Optimize Your Upwork Projects with Make.com

We are building a focused AI receptionist service for independent automotive appearance businesses in the United States, initially targeting PPF, window tint, ceramic coating and vehicle wrap shops.

We are looking for an experienced Voice AI automation specialist to personally build a working, demonstration-ready MVP within seven calendar days.

This is not a request to develop a complete SaaS platform, mobile application or custom CRM. The objective is to create a reliable, realistic and easily replicable AI receptionist that can be demonstrated to prospects and subsequently configured for the first paying pilot customers.

Preferred platform: Retell AI.

Vapi or another platform may be proposed only if you can clearly explain why it is better for this specific MVP in terms of reliability, setup speed, appointment handling, call transfer, SMS and ongoing cost.

Required MVP functionality

The AI receptionist must:

* Answer inbound calls on a US phone number
* Clearly identify itself as an AI assistant for the automotive shop
* Answer basic questions using a business-specific knowledge base
* Provide information about services, business hours and location
* Handle enquiries relating to PPF, window tint, ceramic coating and vehicle wraps
* Collect the caller’s name, telephone number, vehicle make/model/year, requested service and preferred appointment time
* Check calendar availability and book an appointment when appropriate
* Offer a callback request when direct booking is not possible
* Send an SMS or email confirmation to the caller
* Send a structured lead summary to the business owner
* Transfer the call to a human when required
* Handle unknown questions without inventing information
* Record the call outcome, transcript and collected lead data
* Store the lead information in a simple system such as Google Sheets, Airtable or another lightweight solution
* Be easy to duplicate and customize for future pilot customers

The initial MVP may use fictional shop information. We will provide or approve the sample business details, services, working hours, frequently asked questions and escalation rules.

Important limitations

The agent must not:

* Invent prices, discounts, services or available appointment times
* Promise work completion dates
* Give technical advice about vehicle treatment when the information is not included in the knowledge base
* Confirm an appointment without checking actual availability
* continue answering when the situation requires human intervention
* Collect unnecessary sensitive information

A safe fallback response and human escalation process must be included.

Required deliverables

1. A working US demo telephone number
2. Complete inbound Voice AI agent
3. Automotive-specific conversation flow
4. Configured knowledge base
5. Lead qualification and data-capture workflow
6. Google Calendar or Cal.com appointment integration
7. Callback-request workflow
8. SMS or email confirmation workflow
9. Business-owner notification and call-summary workflow
10. Human call-transfer functionality
11. Call recordings, transcripts and outcome tracking
12. Simple lead log in Google Sheets, Airtable or an agreed alternative
13. Export or copy of all prompts, workflows and configurations
14. Written documentation explaining the setup
15. A short recorded walkthrough showing how to customize the system for another automotive shop
16. Full handover of all accounts, configurations and project assets

All main platform accounts must be created under, transferred to or fully controlled by the client. The finished system must not depend on an account that remains exclusively under the freelancer’s ownership.

Testing and acceptance criteria

Before final approval, the system will be tested across at least 30 simulated calls, including:

* New PPF enquiry
* Window-tint enquiry
* Ceramic-coating enquiry
* Vehicle-wrap enquiry
* Price question
* Unknown service
* After-hours call
* Appointment request
* Unavailable appointment time
* Callback request
* Request to speak with a human
* Caller changing details during the conversation
* Question not covered by the knowledge base
* Noisy or unclear caller
* Interrupted or abandoned call

The MVP will be accepted when:

* Required lead information is captured correctly in at least 90% of applicable tests
* The agent does not invent prices, services or appointments
* Calendar bookings and callback requests are recorded correctly
* Notifications and confirmations are delivered correctly
* Human transfers and fallback scenarios work reliably
* Calls, transcripts and outcomes are recorded
* The conversation is natural enough for a real prospect demonstration
* No critical unresolved errors remain

Project structure

Fixed budget: $700
Delivery period: Maximum seven calendar days

Milestone 1 – Architecture and call-flow approval: $100
Final technical stack, conversation map, required integrations and estimated recurring costs.

Milestone 2 – Working MVP: $450
Working US demo number, Voice AI agent, knowledge base, lead capture, calendar/callback workflow, notifications and human transfer.

Milestone 3 – Testing, corrections and complete handover: $150
Acceptance testing, correction of identified issues, documentation, recorded walkthrough and transfer of all project assets.

No paid tool, subscription or external service may be purchased without prior written approval. Please identify all expected third-party and recurring costs in your proposal.

Who we are looking for

You should have direct hands-on experience with:

* Retell AI, Vapi or comparable Voice AI platforms
* Inbound AI receptionists
* Prompt and conversation-flow design
* US telephone-number configuration
* Calendar integrations
* SMS or email automation
* Call transfer and escalation
* Webhooks and API integrations
* Lead capture and lightweight CRM workflows
* Voice-agent testing and debugging

You must personally implement the project. No subcontracting or delegation is permitted without prior written approval.

Application instructions

Please begin your proposal with the words:

“AUTOMOTIVE VOICE MVP”

Then answer all screening questions and include at least one relevant example of a Voice AI agent that you personally built.

Generic AI-generated proposals that do not address the required workflow, technical architecture and delivery timeline will not be considered.

We are a small, growing custom furniture manufacturer looking for an experienced automation developer to streamline our post-order workflow. We currently receive orders through Shopify, Etsy, and Square, use Google Sheets as our internal production board, ShipStation for shipping, and Gmail for personalized customer communication.

We already have established systems and processes that work well. Our goal is not to replace them, but to connect them and eliminate repetitive data entry and manual customer emails.

What We Want to Automate
1. New Order → Google Sheets
When an order is confirmed through Shopify, Etsy, or Square, automatically add the relevant information to our existing Google Sheets production board. A screenshot of our existing production board is attached for reference.
*For custom orders processed through Square, the workflow may begin when the required deposit/payment is received.

Order information includes:
• Customer name
• Order/invoice number and source
• Order date
• Wood species, dimensions, and quantities
• Bracket quantities, finish, and size
• Custom order notes
• Estimated shipping date
• Order/production status

2. Personalized Order Confirmation Email
Once the order/deposit is confirmed, automatically send a personalized thank-you email that includes relevant order information and an estimated shipping date or shipping window based on our current production lead time. (We will provide the email templates and rules for determining estimated shipping dates.)

3. Existing Google Sheets Production Board
Our team will continue using our current Google Sheets board to manage production.
Changes to certain order/status fields should be able to trigger additional automations.

4. ShipStation + Tracking
When an order ships, automatically retrieve the carrier, tracking number, and ship date from ShipStation and associate it with the correct order.

5. Personalized Shipping + Installation Email
Once the order ships, automatically send the customer a personalized email containing:
• Tracking information
• Correct installation instructions based on the products purchased
• Links to the appropriate installation manuals
• Number of wall anchors/hardware required for their specific order using rules/formulas that we provide.
• Any relevant product-specific installation information


Ideal Workflow
Shopify / Etsy / Square → Google Sheets → Personalized Order Confirmation → Production → ShipStation → Personalized Shipping & Installation Email

This is Phase 1 of a larger automation project. If the initial project goes well, we will likely have additional automation work involving invoicing, automated social media, and reporting.

When Applying Please briefly answer the following:
Have you personally built an integration using the Square API or Square webhooks? If yes, please describe it.
Have you integrated ShipStation into an automated workflow?
Would you recommend Make.com, n8n, custom APIs/webhooks, or a combination for this project, and why?
Please share an example of a similar order-management, e-commerce, or operational automation you've built.

We are looking for someone who can build a reliable system that is easy for our team to understand, maintain, and expand over time.

Hi Arifdeen,

Thank you for applying to Audentes Marketing. I reviewed your profile and your experience with Make.com, Voiceflow, and Vapi AI is exceptional.

While this specific job post was designed for high-volume cold outreach and sales, I actually need a lead AI Automation Expert like you to build our core agency templates.

I want to hire you at your $5/hr rate for a specialized technical project. I need you to build a master "24/7 AI Receptionist & Booking Bot" template inside Voiceflow that connects via API to a dummy Google Calendar using Make.com. This template will be used to show local service businesses (like plumbers or roofers) how our agency can automatically capture leads and book their appointments.

Are you available to build this master agency asset for us this week? If so, let me know and I will send over the formal contract offer.

Best,
Khinary Williams
Owner, Audentes Marketing

We need a developer to finish and deploy a Python integration for our lending CRM (GoHighLevel). We have a working starter — Claude reads applicants' tax returns/credit reports and computes a debt/income summary — but it's untested against live GHL, especially the Custom Objects/Associations API. You'd debug those calls, wire up credentials, and deploy on a schedule. Full scope + starter code provided. Python + REST API experience required.

Data implementation from an Excel sheet to an CRM called bullhorn.

Many thanks

Hi,

I'm building AFRILATE, an early-stage friendship platform helping African diaspora adults in London find compatible people based on lifestyle, interests, social style and friendship intentions.

I'm currently looking for a hands-on Meta Ads specialist to run a small proposition-validation experiment, rather than an ongoing social-media-management campaign.

Your background caught my attention, particularly your experience with [INSERT RELEVANT EXPERIENCE FROM THEIR PROFILE].

An essential requirement for this project is genuine experience running paid campaigns targeting African diaspora audiences, rather than simply campaigns targeting people within African countries.

The initial experiment will test three different propositions against the same London African-diaspora audience while keeping the audience, approximate spend, creative quality and landing experience controlled.

Initial Meta media spend will be approximately £100.

The work would involve:

- Meta campaign setup
- London African-diaspora audience strategy
- three controlled proposition tests
- Pixel/conversion-event/tracking verification
- UTM/source attribution
- monitoring and optimisation
- results analysis
- explaining clearly what the small sample does and does not allow us to conclude

This is specifically a validation experiment, not follower growth, content management or general brand awareness.

Before going further, could you answer three quick questions?

1. Can you describe ONE paid campaign you've personally run targeting an African diaspora audience? Please include the diaspora audience, geography, platform, approximate spend and measurable result.

2. How would you approach reaching African diaspora adults in London through Meta without simply relying on "Africa" as an interest?

3. With approximately £100 of initial media spend, what do you think we could reasonably learn, and what could we NOT conclude?

If this is the kind of project you'd be interested in, I'd also be interested in your fixed fee for setting up, running and reporting on the initial experiment.

Thanks,
Toks

Hi Gavin I noticed that you are a Facebook expert I need assistance please. I recently started a Facebook business page but I'm unable to let clients leave a review is that something you can have me with?

We are looking for an experienced n8n automation expert to build a complete content generation and approval system for our marketing department. The goal is to automate the creation of content for Medium, X (Twitter), and our SEO blog, while maintaining a high standard of quality through automated "AI slop" checks and human-in-the-loop approvals.

**Project Overview:**
You will build an n8n or Make.com workflow that connects an LLM (e.g., OpenAI/Anthropic), Notion, and Telegram to manage our entire content pipeline.

**Core Workflow Requirements:**

- **Content Generation:**
- **Medium & SEO Articles:** Automatically generate articles once a week based on predefined topics/niches.
- **X (Twitter) Posts:** Automatically generate posts 3 times a week.
- **Quality Control & "AI Slop" Check:**
- All generated content must pass through a secondary AI prompt designed to detect "AI slop" (overly generic, robotic, or repetitive text).
- **Human-in-the-loop:** If the content is flagged as doubtful or low quality during this check, the workflow must pause and send a Telegram notification to a human reviewer for manual editing/approval.
- **Notion Workspace Integration:**
- Everything must be organized in a Notion database, acting as a Content Calendar with scheduled dates for all posts/articles.
- If an article or tweet *passes* the automated AI check (no human review needed), the system must automatically create a task card in a Notion Kanban board and place it in the **"Ready to Publish"** column. The generated text must be inside the Notion card.
- **Telegram Notifications:**
- When a post is moved to "Ready to Publish," the system must send a Telegram notification.
- The notification must include an inline **"Open"** button containing a hyperlink that takes the user directly to the specific Notion task card.

**Tech Stack You Will Use:**

- n8n (Core automation logic) or Make
- LLM APIs (OpenAI, Anthropic, or similar)
- Notion API
- Telegram Bot API

**Ideal Candidate:**

- Proven experience building complex workflows in n8n.
- Strong understanding of API integrations (specifically Notion and Telegram).
- Experience with LLM prompting and structuring automated AI quality checks.

**To Apply:**
Please start your proposal by sharing a brief description (or screenshots/video) of a similar automation you have built using n8n, especially if it involved Notion, Telegram, or AI content generation. Let me know your estimated timeline for this setup.

We are seeking a proactive automation specialist in n8n and make.com to support our growing marketing division. You’ll create workflows that will work with HubSpot, salesforce, and AI Agent platforms. You will also test and fix workflow bugs. Flexible remote role — ideal for self employees/contractors familiar with automation using n8n and make.com.

Requirements
To qualify, you must meet the following:

✅ Have at least 6 months of prior remote work experience.

✅ Be physically located near the Eastern Time Zone (U.S.) for easier collaboration.

✅ Speak fluent English and communicate confidently on video calls.

✅ Have a reliable internet connection.

✅ Have automation experience (experience with AI agents is a plus).


Responsibilities
You will be responsible for:

Designing and building workflows in n8n and Make.com to facilitate data exchange between Salesforce and HubSpot.

Troubleshooting and optimizing existing workflows to ensure accurate data normalization between CRM platforms (Salesforce and HubSpot) and the AI agent platform.

Collaborating with the Project Manager to identify, define, and implement workflow requirements based on client needs.

I run a boutique travel advisory business helping clients book luxury travel using points and miles. I've built out a working Airtable base that manages client intake (via a Tally form), trip tracking, flight option delivery, and automated client notifications — but I've hit the limits of what I have time to build myself and need an expert to take it further.

What's already in place:

Airtable base with linked tables for Clients, Trips, Flight Options, and Booking Summaries
Multiple live automations (client notifications, status updates, savings calculations)
A Tally intake form feeding directly into Airtable
An early-stage Airtable Interface for a client-facing trip portal

What I need help with:

Polishing and properly designing the client-facing portal (Interfaces or an alternative approach if you have a better one)
Building out and refining automations as the business scales
General Airtable schema/database cleanup and best practices
Ongoing support as new features and workflows come up

Ideal candidate:
Deep hands-on experience with Airtable (Automations, Interfaces, API/scripting)
Comfortable with form tools like Tally (or similar) and connecting them to databases
Bonus: experience designing clean client-facing portals/dashboards
Strong communicator who can work somewhat independently once ramped up

There's potentialy for more ongoing work as well as I have a steady stream of systems/automation projects for this business and am looking for someone to build a long-term working relationship with but we will see how this goes first.

To start: please briefly describe your experience with Airtable Automations and Interfaces specifically or include any relevant portfolio examples.

We are looking for a full redesign of the ReachMyTeach website, not a rebrand. We already have substantial copy, an initial sitemap, product organization, and design ideas in the shared Google Drive folder.

Goals
Create a new, intuitive sitemap
Clearly explain the full ReachMyTeach platform
Make a complex product feel simple and approachable
Build a polished, responsive WordPress site
Make the site easy for our team to edit without coding
Use reusable templates and sections to keep the design consistent
Preserve important existing content and URLs
Complete the work on a staging site before launch
Key Pages
Homepage
Five solution hubs
Priority product and feature pages
Plans and pricing
About
Resources and customer stories
Contact and Request a Demo

White-Label AI Sales & Outreach Automation Platform

We are looking for an experienced AI sales automation specialist or small agency to design and build an intelligent, end-to-end sales and outreach system that we can white-label and deploy for multiple clients.

This is not simply a GoHighLevel setup or an Instantly cold-email campaign.

We want to create a complete AI-powered sales engine covering:

Identify → Research → Enrich → Prioritise → Personalise → Contact → Engage → Qualify → Book → Nurture → Measure → Optimise

The system should include:

Automated prospect/data identification and enrichment
AI research of companies and decision-makers
Highly personalised, natural outreach at scale
Multi-channel engagement across email, LinkedIn, SMS, WhatsApp and potentially voice
Intelligent sequencing based on prospect behaviour and responses
AI agents capable of handling replies, qualification and appropriate follow-up
GoHighLevel or equivalent CRM and pipeline management
Automated calendar/appointment booking
Human handover when appropriate
Client-facing dashboard and analytics
Repeatable white-label architecture for rapid deployment to new clients

Our initial thinking includes GoHighLevel, Instantly, Clay, n8n/Make and AI agents, but we are not committed to this architecture. We want an expert who can recommend a better stack where appropriate.

A useful benchmark for the type of intelligent GTM automation we are interested in is Alta (altahq.com). We are not asking for a copy of Alta, but the principle of combining data, research, personalisation, multi-channel outreach and AI-driven engagement is relevant.

This is intended to become an ongoing relationship

The initial requirement is to design and build the system and prove it with an initial deployment.

We then intend to sell the solution to multiple clients.

We therefore particularly want to hear from suppliers who can provide ongoing client fulfilment, including onboarding, configuration, campaign setup, data sourcing, deliverability, optimisation and technical support.

This could become substantial recurring work for the right supplier.

When applying

Please read the attached detailed brief before responding.

We are particularly interested in applicants who have already built comparable end-to-end systems.

Please include:

Examples of similar systems you have built.
The technology stack you would recommend and why.
How you would approach the white-label/multi-client architecture.
Which parts of ongoing client fulfilment you could provide.
Your hourly rate and an approximate time and total cost for the initial build.

Please do not send a generic AI or GoHighLevel proposal. We are specifically looking for someone with experience combining sales automation, AI agents, outbound infrastructure, CRM, data/enrichment and multi-channel engagement.

I’m building a company that helps service businesses modernize their operations using existing tools and automations.

I handle sales, discovery calls, client communication and project management. I’m looking for a reliable technical specialist who can handle implementation work such as CRM setup, n8n/Make/Zapier workflows, lead routing, SMS/email automation, AI voice, APIs and integrations.

This is not a full-time role right now. I’m looking for someone I can work with on a per-project basis and potentially send ongoing client work to.

I’d like to start by reviewing your past work and having a quick conversation. If it looks like a fit, I’ll hire for a small paid trial project.

Please include:

* examples of systems you’ve built
* tools you’re strongest with
* typical project pricing
* availability
* how you handle testing and failures

I’m looking for a ClickUp expert to make a small improvement to an existing CRM and client management workflow.

The task includes:

Review one ClickUp CRM/List
Improve the client tracking structure
Add or adjust relevant statuses or Custom Fields
Configure one simple client-management automation
Test the workflow with a sample task/client

Example:

When a lead becomes a client → update the status, assign the account owner, and move the item into the client workflow.

This is a small fixed-price task.

Programmatic Local SEO Specialist – Build Dynamic Multi-City Service Matrix (150 Pages) in WordPress

Job Overview / Scope of Work:
We are a residential property services and exterior maintenance platform launching a new digital footprint in a major US metro market.

We are hiring an experienced WordPress & Local SEO developer for a fixed-price project to design and deploy a programmatic local SEO landing page matrix covering 15 regional municipal zones across 10 service categories (150 total programmatic location pages).

(Note: Full brand domain, exact city dataset, and detailed service copy will be provided directly to the selected candidate upon hire).

Core Objectives & Deliverables:

Master Landing Page Template Design:
Design 1 modern, mobile-first, high-converting service template.
Key conversion elements: Sticky mobile call/quote button, clear hero section, dynamic service selector, quote request / photo upload form, review proof section, and dynamic FAQ accordion.

Programmatic Generation (150 Dynamic Location Pages):
Programmatically generate 150 landing pages (15 regional target cities × 10 service verticals) using a dynamic tool (such as Page Generator Pro, WP All Import, or custom Gutenberg blocks).
Clean, SEO-optimized URL structure (e.g., brand.com/service-slug-city-slug-fl/).

10 Service Categories (Dataset provided on onboarding)

Local Schema Markup (JSON-LD) & Dynamic SEO:
Configure RankMath SEO (or similar) to inject dynamic Local Business Schema (HomeAndConstructionBusiness) incorporating geo-coordinates, city names, and service types.
Dynamic meta titles, open graph tags, and meta descriptions optimized for local click-through rate.

Performance & Integration Readiness:
Performance optimization for sub-second mobile page loads on a LiteSpeed Web Server.
Webhook integration on quote forms (Airtable /Make.com ready).

Requirements:
Proven portfolio building programmatic local SEO sites or multi-location directories in WordPress.
Deep knowledge of Local Schema (JSON-LD) and on-page technical optimization.
Experience with lightweight themes (Astra, GeneratePress, or Bricks) and LiteSpeed performance caching.

To Apply:
Please share links to 2–3 live programmatic local SEO sites you have personally built.
Briefly state the specific plugin/tool you prefer to use for dynamic mass-page generation in WordPress.
Confirm turnaround time

We are looking for a freelancer to quickly finalize a WhatsApp reminder automation connected to our Google Workspace Calendar.

We already have:

* Verified Meta Business account
* Verified WhatsApp phone number
* Active Make.com account
* Existing Make workflow that needs to be fixed/improved
* Approved WhatsApp message templates already created
* Google Workspace / Google Calendar
* Notion database containing our contacts and phone numbers

Required workflow:

Google Calendar appointment → retrieve attendee email → search the email/contact in Notion → retrieve the phone number → automatically send WhatsApp reminders:

* 24 hours before the appointment
* 3 hours before the appointment

The current workflow already exists but is not working correctly. We are looking for someone who can troubleshoot it, fix it and modify/rebuild parts of it if necessary.

We are also open to using other tools/providers if they make the setup more reliable or easier to manage, such as Twilio, Zapier, other WhatsApp API providers or SaaS platforms.

Nice to have: a simple solution that also allows our team to receive, read and potentially reply to incoming WhatsApp messages from the same number.

Goal: have the system fully working within the next 3–4 days.

Budget: approximately €350 for the initial setup.

Please apply only if you have direct experience with Make.com, WhatsApp Business API / Meta Cloud API, Google Calendar and API/webhook automations.

In your application, briefly explain:

1. Similar WhatsApp automations you have built
2. Which architecture/tools you would recommend
3. Whether you can deliver a working solution within 3–4 days

We’re looking for an experienced GoHighLevel specialist to help us set up, connect, and optimize our GHL Forms, Workflows, and CRM. The goal is to create a smooth and reliable system where leads are captured, organized, managed, and automatically followed up with minimal manual work.

You should be comfortable working with GoHighLevel workflows, CRM automation, forms, triggers, custom fields, integrations, and lead management systems. Experience troubleshooting and improving existing GHL setups is a plus.

Scope of Work

* Set up and optimize GoHighLevel workflows
* Connect and configure GHL Forms with the CRM
* Create triggers and automated follow-up sequences
* Configure custom fields and lead organization
* Set up lead capture and management processes
* Connect GHL with external platforms when required
* Troubleshoot existing workflows and automations
* Test the complete automation process and ensure everything works correctly
* Improve the overall efficiency of our CRM and lead management system

Requirements

* Strong experience with GoHighLevel/HighLevel
* Experience with CRM and marketing automation
* Knowledge of workflows, triggers, forms, custom fields, and pipelines
* Experience with API and third-party integrations
* Familiarity with n8n, Make.com, or Zapier is a plus
* Strong troubleshooting and problem-solving skills
* Ability to follow instructions and document completed work

We're a lead generation company. We push over $1M a month of ad spend across Meta, Google, TikTok and LinkedIn, and route the leads that come back to contractor clients in real time, against each client's coverage area, remaining budget and daily caps. Around 100 automated jobs run all of that, plus ZIP code targeting, budget rules, invoicing, CRM syncs and reporting. Nobody watches them.

Here's our problem. They almost never crash. They break quietly, keep reporting success, and we find out weeks later because somebody noticed a number looked off.

One example. A daily job stopped fetching one of its three levels of ad spend data for 28 days. The other two kept working, so the dashboard still filled in and nothing errored anywhere. $372,000 of ad spend ended up attributed to the wrong campaigns. It passed our tests, our linting and our code review.

We want someone to own this end to end:

- Find what isn't doing its job, including the things currently reporting success
- Fix them, and leave a test that fails if someone later undoes the fix
- Build detection for every automation we run. Error alerts catch none of these, because nothing errors.
- Verify before shipping: tests you've confirmed can actually fail, dry runs, output checked against the source of truth
- Build new automations to the same bar, and tell us which of the existing ones should be deleted rather than fixed

There's a technical context document attached with our full stack, how the code is structured, and five more real failures. Worth reading before you apply.

You're a good fit if:

- You've been responsible for something considerably more complex than this, ideally where a quiet failure cost real money
- You reproduce a bug before you fix it, and you check the real numbers against the source instead of assuming the code did what it says
- You'd rather ship one thing you've verified than three that look right
- You use AI coding tools and can tell us exactly how you check their work

You're not a fit if:

- You'd ship what the model gave you because it ran and the output looked right
- You'd describe your main strength as speed
- You're an agency, or you'd hand this off to someone else

What you'd be working in. The automations are roughly 80% Cloudflare Workers and 20% Make.com and n8n, and the Make and n8n ones run critical parts of the business. They sit across a codebase of about 90,000 lines covering three products that are all still evolving: a CRM, a dashboard app, and the operations hub the company runs on. Our databases are in the low millions of records. If you haven't worked in a codebase that size, or with data at that scale, don't apply.

Stack: JavaScript on Node, Cloudflare Workers, Postgres and Supabase, Make.com, GitHub Actions, and about a dozen third party APIs including Meta, Google Ads, Stripe, Airtable and GoHighLevel.

Requirements: individual only, no agencies and no subcontracting. Most of your working day overlapping 9 to 5 US Eastern. Willing to fix something urgent in the evening a few times a month. Fluent English. Full time, 40 hours a week, long term.

To apply, answer these four questions. We read the answers, not the cover letter.

1. What's the most complex system you've been responsible for (scale, stakes, cost of a quiet failure)?
2. Tell us about an automation or data pipeline that was broken for weeks before anyone noticed. How did you find it, and what did you change so that kind of thing couldn't hide again?
3. Walk us through the last non-trivial thing you shipped, from finishing the code to it being live. What did you actually do, not what you'd do ideally.
4. How do you use AI coding assistants, and how do you check their work before it goes out?

We are looking for an experienced CRM Automation Developer to set up and optimize our sales and lead management workflow.

The work will include CRM setup/customization, lead capture and automatic assignment, sales pipeline automation, follow-up reminders, email/WhatsApp automation, status updates, reporting dashboards, and integration with external tools or APIs.

The developer should be able to understand our current business process and convert it into a clean, reliable automated CRM workflow with proper testing and documentation.

Experience with platforms such as Zoho CRM, HubSpot, GoHighLevel, n8n, Make.com, Zapier, APIs, and webhooks will be preferred.

Please share 2–3 relevant CRM automation projects or examples with your proposal.

We are looking for someone who can also support future CRM and automation requirements if the initial project goes wel

Skills: Software Architecture, CRM, Lead Generation, Automation, Hubspot, Zapier, API Integration, Zoho CRM, n8n, Workflow Automation

We are launching StoryTours.ai, an editorial digital audio publisher producing narrative-driven historical walkthroughs of premier European landmarks (launching initially with Versailles, expanding to London and Paris).

Our brand tagline is: "History is theater. Take the best seat."

Crucial Architecture Note: Our fulfillment backend is already 100% built and operational (Stripe webhooks to Make.com to Airtable to Postmark to VoiceMap). You do not need to build backends, databases, or automation scenarios.

Your sole responsibility is the high-performance front-end static site, comprehensive machine-readable schemas, ensuring our tour purchase buttons pass our live inventory parameters cleanly through Stripe Payment Links, implementing a Markdown-based seasonal tour availability system, and structuring an extensible multi-language setup (launching in English, with French, Spanish, German, and further locales added shortly after).

We launch initially in English with three definitive Versailles productions:

The Ritual of Power: Unofficial Versailles State Rooms Guide (/en/versailles-state-rooms) — Price: $9.99 USD | Duration: 60 MIN | SKU: RITUALS-English (Year-round Palace interior)

The Architecture of Control: A Guide to the Versailles Gardens (/en/versailles-gardens) — Price: $8.99 USD | Duration: 90 MIN | SKU: ARCHITECTURE-English (Seasonal: Active April 1 – October 31; Closed November 1 – March 31 due to locked grove gates)

The Illusion of Innocence: Queen's Hamlet & the French Revolution (/en/queens-hamlet) — Price: $8.99 USD | Duration: 60 MIN | SKU: MARIE-ANTOIN-English (Year-round Queen's Estate)

We need a lightweight static site built for two distinct users:

Tourists on Mobile: Standing in museum queues on congested cellular networks needing sub-1.5s page loads, zero audio-streaming layout lag, frictionless one-tap checkouts, and cookie-free analytics.

AI Engines & Web Crawlers: LLM scrapers (Gemini, ChatGPT, SearchGPT) requiring semantic HTML, structured JSON-LD schemas, root llms.txt, and public machine-readable endpoints.

Scope of Work

Front-End Core, Compliance, & Content System

Framework: Static Site Generation (SSG) via Astro (strongly preferred for near-zero JS overhead) or lightweight Next.js.

Extensible i18n Routing Structure: Astro built-in i18n routing configured with defaultLocale: 'en' and pre-configured support for locales: ['en', 'fr', 'es', 'de'] plus future additions. Root path / redirects cleanly to /en/. Dynamic canonical tags and alternate language tags injected automatically into every page header across all active locales. A clean, extensible language switcher in the header nav.

Performance & Analytics Budget: 95+ Google PageSpeed Mobile score across all pages. Initial page payload strictly under 1.5 MB. Integrate a lightweight, privacy-first analytics script (e.g., Plausible or Fathom) that respects the payload limit and avoids GDPR cookie banners.

Audio Player Engineering: Embedded HTML5 preview players must strictly use preload="none" to prevent background cellular data consumption. Audio streams asynchronously only upon user interaction.

Scalable Multi-Language Markdown Schema: All tour content and legal pages must live in structured Astro Content Collections organized by locale subdirectories: src/content/tours/[locale]/[slug].md (e.g., /en/, /fr/, /es/, /de/). Adding a translated tour (or a whole new language) must require zero template or routing edits—only adding a Markdown file. Frontmatter includes: locale, title, sku, price, duration_minutes, is_active, seasonal_notice, audio_preview, stripe_payment_link.

Legal & Compliance Pages: Standard static footer routing to /privacy, /terms, and /refunds across all active locales, rendering from Markdown files.

Seasonal Availability & Winter Lockout Engine: Support is_active: boolean and seasonal_notice: string in Markdown frontmatter. When is_active: false (e.g., winter grove closures from Nov 1 to Mar 31), the Stripe "Buy" button is completely disabled and replaced with (Closed for Season — Reopens April 1), displaying an inline fallback banner pointing visitors to the open State Rooms tour and providing an email waitlist field.

Homepage Experience: Editorial hero section with brand framing and an embedded audio prologue player. Instant client-side destination filter bar: (Versailles Live) (London Coming Soon) (Paris In Production). Unreleased destinations surface a clean inline email capture form. Tour cards with metadata badges (60 MIN | PALACE INTERIORS, 90 MIN | GARDENS & GROVES, 60 MIN | QUEEN'S ESTATE) and single-click checkout buttons.

Dedicated Tour Landing Pages: Build the three launch routes under /en/ (with templates dynamically handling translations). Includes hero block, 30-45s async audio teaser, 3-4 visual stop cards with captions, logistical badges, Google Maps meeting pin, and a localized /success confirmation page with 3-step VoiceMap redemption instructions.

Expected Billing Experience & Stripe Parameter Wiring

The Buyer Journey: Purchase buttons open pre-configured Stripe Payment Links prioritizing Apple Pay and Google Pay. Zero billing friction: email only (no shipping/billing addresses or phone fields). Multi-ticket checkout enabled (1 to 20 vouchers). Direct redirect to the matching localized success page post-payment.

Technical Parameter Passing & Test Mode: Buttons append tracking parameters via client_reference_id passing match keys (e.g., RITUALS-English, and dynamically RITUALS-French, RITUALS-Spanish, RITUALS-German as translated files are added) alongside quantity and language. Testing will be conducted using Stripe Test Mode links.

Machine-Actionable SEO, E-E-A-T & LLM Standards

Root LLM Standards: Deploy /llms.txt and /llms-full.txt specifying boundaries, operating seasons, runtimes, available and planned languages, pricing, and checkout endpoints.

Granular JSON-LD Schemas: Deploy nested Schema.org microdata (Product & Offer with inLanguage, dynamic InStock/OutOfStock availability, AudioObject with PT60M/PT90M durations, TouristAttraction, Person, and FAQPage). (Do NOT include AggregateRating).

Agent Query Endpoints: Public /openapi.json spec and static /api/v1/tours.json catalog endpoint.

Deliverables & Milestones (Total Fixed Price: $600)

Milestone 1: Static Architecture, Editorial Front-End, i18n & Legal Pages ($250)

GitHub repository setup connected to Cloudflare Pages or Vercel.

Astro Content Collections configured with scalable language directories (/en/, /fr/, /es/, /de/) and frontmatter schemas.

Instant destination filter bar, tour cards, responsive mobile layout, extensible language switcher, and standard footer with /privacy, /terms, /refunds.

Dedicated landing pages built for the three Versailles launch titles in English (including inactive winter state styling for Gardens).

Localized post-checkout confirmation page (/en/success).

Milestone 2: Machine-Actionable Data, Schema, SEO & Analytics ($150)

Validated JSON-LD schema suite tested against Google Rich Results.

Dynamic canonical tags, localized alternate language tags mapped across all configured locales, XML sitemap via @astrojs/sitemap, OpenGraph tags, root /llms.txt, /llms-full.txt, and static /api/v1/tours.json.

Implementation of Plausible/Fathom analytics script.

Milestone 3: Stripe Checkout Wiring, Test Mode Verification & Handover ($200)

Stripe buy buttons connected with parameter passing (client_reference_id with dynamic SKU language handling) verified using Test Mode links.

Confirmation of frictionless mobile checkout flow (Apple Pay trigger to redirect to /success).

Production launch on custom domain (storytours.ai) swapping in Live Stripe links.

5-minute screen recording demonstrating:

Toggling seasonal tours on/off.

Adding new translated tours (e.g., /fr/, /es/, /de/) or completely new locales without touching code.

Adding new cities (London, Paris) via simple Markdown files.

Editing legal/refund pages.

Application Screening Questions (Please answer when submitting your proposal):

Share 2 examples of production websites you built using Astro with native multi-language (i18n) routing that achieve a 95+ score on Google PageSpeed Mobile.

How do you pass custom parameters (tour_id, quantity, language) through a Stripe Payment Link URL so the downstream webhook receives them in client_reference_id?

How do you structure Astro dynamic routes ([locale]/[slug].astro) so adding a new language folder (/es/, /de/) automatically generates pages with correct language tags without touching code?

Confirm you will provide a brief video handover showing how to toggle seasonal tours and add translated content via Markdown.

Vision Rescue is a non-profit organisation serving children from urban slum communities to drive inter-generational transformation. We operate across Maharashtra, Tamil Nadu, Telangana and West Bengal.

This year we are building systems for local fundraising and are looking for a developer to set up a clean, ready-to-use workflow. On completion, we would like to explore a one-year hourly contract for consultancy and maintenance.

Please refer to the brief below for further details on the project.

We are looking for an experienced Outgrow platform expert to assist with the completion, troubleshooting, and optimization of a complex, multi-branching diagnostic survey and lead-generation assessment. Both the questions and the results sections of the assessment have already been created.

The project involves creating complicated conditional logic paths based on user roles and multi-select method triggers in addition to constructing dynamic outcome logic and producing PDF reports automatically.

Project overview

First Sign Solutions Ltd is a UK healthcare solutions business exploring a potential recruitment workflow automation project for a healthcare provider.

We are looking for an experienced automation specialist to advise on and potentially build an MVP that streamlines candidate intake, qualification, administration and interview scheduling.

Proposed workflow:

Candidate enquiry/application → structured qualification form → CV upload → extraction/summarisation of relevant CV information → CRM/database record → rules-based qualification/routing → human review → interview booking → automated reminders/follow-up → simple reporting dashboard.

AI may be used to assist with extracting and summarising information from CVs, but AI should not autonomously make hiring or rejection decisions. Final recruitment decisions must remain with humans.

We are open to recommendations on the technology stack, but relevant experience may include:

• Make.com and/or n8n
• Airtable, HubSpot or similar CRM/database
• Jotform, Tally or similar forms
• OpenAI/other suitable LLM APIs
• Email and calendar integrations
• Automated notifications and follow-ups
• Reporting/dashboard creation

Important requirements

Because this is intended for a UK healthcare environment, we would particularly value someone who understands data privacy, UK GDPR principles, secure handling of candidate information, access controls and data minimisation.

The solution should ideally be built within accounts owned by us/the client rather than the freelancer, with appropriate documentation and handover.

At this stage, we are primarily assessing technical feasibility, architecture, implementation time and cost before proceeding with a build. No real patient or candidate data will be provided during initial development/testing.

When applying, please answer:

1. What technology stack would you recommend for this workflow and why?
2. Would you recommend Make.com or n8n for this use case?
3. How would you handle CV data extraction and summarisation?
4. How would you approach UK GDPR/privacy and security?
5. How would you ensure human review remains part of candidate decision-making?
6. How would you monitor and handle automation failures/errors?
7. Can the complete workflow be built in accounts owned by us/the client and handed over with documentation?
8. Approximately how long would an MVP take?
9. Please provide an estimated fixed price for the MVP.
10. What ongoing software and maintenance costs would you expect?

Please also share examples of similar recruitment, CRM, intake or workflow automation projects you have previously completed.

We are looking for a practical, maintainable no-code/low-code solution rather than an unnecessarily complex custom software build.

Estoy buscando un especialista en HubSpot para ejecutar técnicamente un proyecto de implementación CRM para un cliente.

El proyecto incluye:

configuración y arquitectura de HubSpot CRM;
Companies / Contacts / Deals y asociaciones;
pipelines y propiedades personalizadas;
formularios y landing pages;
tracking mediante UTM y códigos QR;
atribución de campañas y fuentes;
atribución de referrals a empresas/contactos;
automatizaciones y notificaciones de nuevos leads;
importación y limpieza de prospectos;
dashboards y reporting;
pruebas end-to-end;
documentación y vídeo de handoff.

Busco a alguien que pueda encargarse personalmente de la implementación técnica, mientras yo coordino el proyecto, QA y la relación con el cliente.

El plazo sería de aproximadamente 5–7 días.

Antes de avanzar, me gustaría saber:

¿Has realizado personalmente implementaciones completas de HubSpot similares?
¿Tienes experiencia con atribución de campañas/referrals y reporting de revenue?
¿Puedes encargarte también del testing y documentación?
¿Cuántas horas estimas que necesitarías para este alcance?
¿Aceptarías un proyecto cerrado de 500–600 USD si definimos exactamente el alcance antes de comenzar?
¿Cuándo podrías empezar?

Si te interesa, puedo enviarte el briefing completo para que valores el alcance con precisión.

Gracias.

We are looking for an experienced Zoho CRM and QuickBooks Online integration specialist to perform comprehensive system and end-to-end testing of an existing integration between Zoho CRM and QuickBooks Online.

The integration has already been configured/developed, and the primary goal of this project is to validate that data moves accurately and reliably between the two platforms across all expected business workflows.

This is not simply a UI testing assignment. We need someone who understands how CRM records, customers, products/services, estimates, invoices, payments, accounting data, IDs, status changes, and synchronization logic should behave across Zoho CRM and QuickBooks Online.

The selected freelancer will test normal workflows, edge cases, failures, duplicate prevention, field mappings, and synchronization behavior, document any issues found, and work with our team to confirm fixes.

- Responsibilities
You will be responsible for testing the full integration workflow between Zoho CRM and QuickBooks Online, including:

Review the existing integration architecture, workflow, field mappings, and synchronization rules.
Understand which system acts as the source of truth for each type of record.
Validate customer/contact synchronization between Zoho CRM and QBO.
Verify correct creation and updating of customers in QuickBooks from Zoho records.
Test company, contact, billing, shipping, email, phone, and other mapped customer fields.
Validate synchronization of products, services, SKUs, pricing, descriptions, tax settings, and related information where applicable.
Test estimate/quote creation and synchronization.
Test invoice creation from Zoho CRM into QuickBooks Online.
Verify invoice line items, quantities, rates, discounts, taxes, totals, customer assignments, dates, terms, and memo/reference fields.
Validate updates to existing invoices and determine whether changes are correctly synchronized.
Test payment and invoice-status synchronization where included in the integration.
Validate paid, partially paid, overdue, voided, canceled, and outstanding invoice scenarios.
Confirm that QuickBooks IDs and Zoho record IDs are correctly stored and associated to prevent duplicate records.
Test duplicate detection and duplicate-prevention logic.
Test one-way and two-way synchronization workflows where applicable.
Validate synchronization when existing records are already present in one or both systems.
Test record updates after the initial synchronization.
Verify how deleted, inactive, archived, or merged records are handled.
Test required fields, null values, malformed values, missing records, and incomplete data.
Test API and integration failures, including authentication problems, expired tokens, rate limits, network interruptions, and invalid payloads where practical.
Confirm failed transactions are logged appropriately and can be retried without producing duplicates.
Validate scheduled or automated synchronization processes.
Test manual synchronization or retry functionality where available.
Review integration logs, API responses, automation history, webhook activity, or middleware execution history to identify root causes of failures.
Confirm that errors are visible enough for administrators to diagnose and resolve.
Perform regression testing after identified issues are corrected.
Key Test Scenarios

Testing should include both successful workflows and edge cases.

Typical scenarios may include:
New Zoho customer creates the correct QBO customer.
Existing QBO customer is correctly matched rather than duplicated.
Updates to customer information synchronize correctly.
A Zoho quote/estimate creates the appropriate QuickBooks transaction.
A Zoho deal or approved workflow generates the correct QBO invoice.
Multi-line invoices retain the correct products, quantities, pricing, discounts, and taxes.
Duplicate sync attempts do not create duplicate customers or invoices.
QuickBooks payment activity updates the appropriate Zoho record.
Partial payments are represented correctly.
Failed transactions can be retried safely.
Missing or invalid accounting data produces a controlled error rather than corrupting records.
Existing records are matched using the intended IDs or matching rules.
Record edits after synchronization behave according to the defined business rules.
Simultaneous or closely timed updates do not cause inconsistent records.
Authentication or API errors are correctly logged and recoverable.

Additional scenarios should be identified after reviewing the actual implementation.

- Data Validation
A major part of this project will be validating data consistency between the two systems.

We expect the tester to compare:
Zoho record IDs
QuickBooks entity/transaction IDs
Customer/company names
Contact information
Billing and shipping information
Products and services
SKUs/item references
Quantities
Unit prices
Discounts
Taxes
Subtotals
Invoice totals
Payment amounts
Balances
Transaction dates
Due dates
Payment terms
Invoice/estimate numbers
Status fields
Custom fields
Sync timestamps
Error and retry statuses

Financial totals should be verified carefully because even small mapping or rounding issues can create accounting inconsistencies.

- Deliverables
At the end of the project, we expect:
A structured test plan covering all major integration workflows.
Test cases with expected and actual results.
Documentation of all identified bugs and inconsistencies.
Clear reproduction steps for each issue.
Screenshots, logs, record IDs, or API information where useful.
Classification of issues by severity and business impact.
Identification of whether an issue originates in Zoho CRM, QuickBooks Online, the API/integration layer, automation logic, or data configuration.
Recommendations for resolving integration issues.
Regression testing of corrected issues.
Final testing summary showing passed, failed, and unresolved scenarios.
Confirmation of whether the integration is ready for production use.
Required Experience

The ideal candidate should have strong hands-on experience with:
Zoho CRM
QuickBooks Online
Zoho CRM APIs
QuickBooks Online / Intuit APIs
REST APIs
OAuth 2.0
Webhooks
CRM and accounting system integrations
Data mapping and transformation
Integration testing
API testing
System testing
End-to-end workflow testing
Troubleshooting synchronization issues
Duplicate prevention and record matching
Error handling and retry logic

Experience with Zoho Flow, Zoho Deluge, Make.com, Zapier, custom middleware, or similar integration platforms is helpful depending on how the current integration was implemented.

- Ideal Candidate
We are looking for someone who can think beyond individual test cases and understand the complete business workflow.

You should be comfortable tracing a transaction from Zoho CRM through the integration layer into QuickBooks Online and determining exactly where a problem occurred if the final accounting record is incorrect.

Strong attention to detail is important because this integration involves financial information, and incorrect customer matching, duplicate invoices, incorrect line items, tax differences, or synchronization errors can create significant accounting problems.

The ability to clearly document technical issues for developers and explain their business impact to non-technical stakeholders is also important.

- Project Goal
The goal is to ensure that the Zoho CRM + QuickBooks Online integration is stable, accurate, duplicate-safe, recoverable, and ready for production use before relying on it for day-to-day business operations.

Please include examples of previous Zoho CRM and QuickBooks Online integrations or testing projects you have handled, particularly projects involving customer synchronization, estimates, invoices, payments, APIs, or accounting workflows.

We are looking for a Dynamic Email marketer who can jump in for a focused 15-day push and run both our email and SMS promotions end-to-end. The immediate objective is simple and measurable: boost community engagement around an upcoming event while keeping list health pristine—zero spam complaints, clean bounces, full compliance.

Here’s the flow you’ll own:
• Set up segmented SMS sends, email automation campaign .schedule them intelligently across time zones, and continually A/B test copy, timing, and CTAs. (Build a Database of 5k)
• Build and deploy email sequences—design the templates, schedule dispatch, then track open and click-through rates so we can quickly iterate.
• Locate 10-15 relevant LinkedIn, WhatsApp, Facebook, Slack, or Discord communities and tailor announcements for each platform to drive conversations back to us.
• Coordinate with our founders and scheduled speakers so content rolls out on cue.
• Keep the database spotless: handle unsubscribes, hard/soft bounces, and spam-filter best practices throughout.

Tools you can pick from include Mailchimp, HubSpot, Make.com on the email side and Twilio or Klaviyo for SMS; if you have an equivalent stack that achieves the same results, I’m open to it.

Deliverables over the 15-day engagement:
1. A refreshed, properly segmented list of at least 5 k engaged contacts.
2. Two fully automated email sequences and three SMS touchpoints, each with performance reports.
3. Proof of posts in 10-15 community groups with platform-specific copy.
4. Final hygiene report showing list growth and zero spam complaints.

If you’ve run high-velocity promo cycles before and thrive on data-driven tweaks, let’s talk.

Skills: Internet Marketing, Bulk Marketing, Leads, Email Marketing, Twilio, A/B Testing, Hubspot Marketing, Community Management

I’m looking for someone experienced with CRM setup, outbound email campaigns, and lead generation. You should be comfortable working with tools like Clay, Apollo, and Instantly.

I’m looking for someone reliable who can help set things up properly, keep everything organized, and improve the process over time. Good communication and attention to detail are important.

READ THIS FIRST

This is not a "set up a board and a few automations" job. It is the complete operating system for a subscription service — designed in detail over several weeks and documented column by column — that has to run for years without anyone remembering anything. We are looking for one person (or a small team) who can own the entire Monday.com build plus the Make (make.com) integrations end to end, working from a written specification, and prove it works with a documented acceptance test.

If you have not personally built a Monday.com system with 50+ automations, multi-board relationships, date-offset logic, and Make scenarios that write data back into Monday, please do not apply. If you have, read on — this is a substantial, well-defined project with a single decision-maker and a finished design.

1. WHO WE ARE

G. Batista Engineering & Construction (GBE&C) is a structural engineering and construction firm in Fort Lauderdale, Florida. For 26 years we have done building recertification inspections, Structural Integrity Reserve Studies (SIRS), concrete restoration, and owner's-representative work for condominium associations.

Monday.com has been our operational hub for about four years. Today we run six service-line boards with roughly 130 automations, DocuGen for document generation, SuperMail for templated email, a DocuSign integration that drives proposal status, and a Webhooks → Make → QuickBooks Online pipeline that creates invoices and writes payment status and invoice numbers back to the board. We are mid-way through a migration from monday Work Management to monday CRM with our existing consultant; that migration is not part of this job, but what you build must fit the CRM structure we are moving into.

2. WHAT WE ARE BUILDING

SHIELD is a new annual membership program for Florida condominium associations. A building joins, pays an annual fee, and for the life of the membership receives a defined stack of services: written notices ahead of every statutory deadline (some of which carry a contractual guarantee that depends on the notice having been sent and stamped), a monthly written update, an annual inspection walk with a condition report card, a budget-season package tied to the association's own budget-adoption date, insurance-renewal support, event-driven deliverables (bid reviews, repair closeouts, budget alignment letters), and — after a vesting period measured in paid months — the building's next recertification included with membership. A property-manager channel rides on the same machine.

The entire program is delivered through Monday.com. The governing design rule is:

Monday owns WHEN and WHO. An AI drafting layer owns WHAT (first drafts). People own SEAL and SEND.

Every date, alarm, payment status, and entitlement counter must be a deterministic Monday automation. Nothing that carries a guarantee may ever depend on an AI job, a third-party scheduler, or a person remembering. Nothing client-facing leaves without a human approval status somewhere upstream.

3. WHY THIS IS ON THE COMPLEX SIDE

We want you to see the size of this before you quote. Ballpark counts, to be validated by you in Phase 0: four boards plus Monday's hidden subitem board, 120–160 columns in total, three forms, roughly 100–150 native automations, 5–6 DocuGen templates, 15–20 email templates, and 5–7 Make scenarios. The specific things that make it hard:

1. A date-offset engine on connected dates. Every statutory deadline on the calendar board generates three written notices at 12, 6, and 3 months out. Each is sent by the platform the moment its date arrives, to a contact held on the parent board, and stamped on the calendar row with date and time. If the stamp is missing two business days later, two named people are alerted. Budget season runs entirely on offsets from a per-building anchor date (day −300, −120, −105, −90, −30, and +45), so a building with a June fiscal year and a building with a calendar fiscal year run the same automations on different days.

2. Three different spawn events with three different triggers. (a) "GO" — payment-driven: when the first invoice is paid, Monday creates roughly ten Year-1 deliverable subitems under the building's row, twelve monthly-update subitems with due dates, and five starter rows on a second board with the dates deliberately blank. (b) The anniversary — date-driven, not payment-driven: on the date, next year's deliverables and twelve new monthly subitems are created, and recurring calendar rows roll forward one year with a different status ("Confirm date," never "Date needed"). (c) An event form — one button on the building's row opens one form; depending on the event type chosen, Monday creates one to three different objects on two or three different boards.

3. A full lifecycle state machine with human gates. Groups are stages: New Proposal → Proposal Sent → Contract Signed / Onboarding → Active → Suspended → Exiting → Terminated/Alumni — and a building that comes back is reactivated on the same row, never a new one, with its counters reset and its history intact. Money states: delinquent at day 31; a suspension that is a human's click on a platform-drafted notice, never automatic; a 10-day cure; a good-standing flag that gates some sends but must never gate others; a termination click after 60 days; an exit date computed as the last day of the month following the notice month; cancellation (never deletion) of every deliverable due after the exit date; and one final handoff deliverable spawned on exit.

4. A monthly issue with a hard hold. A small control board (one row per month) holds two approval statuses. Every member building has its own pre-spawned subitem for that month that a person must mark "Checked." The send fires only where the building is in good standing AND the subitem is Checked AND both approvals are set; sent / held / skipped counts are written back to the control row; there is no fallback text, ever — the platform nags the right person instead.

5. Money through Make → QuickBooks Online. An annual invoice whose onboarding-fee line is shown or waived depending on the payment election; a quarterly variant; a renewal invoice at day −60 with an escalator; a recertification invoice whose amount is computed from a tenure counter at the moment of delivery; a rejoin invoice; a credit memo on one narrow path triggered by a button; payment status and invoice number write-back; and a Paid Months counter that only ever moves when an invoice is actually paid.

6. An inspector's phone form feeding a per-year board. Ten questions, one branch, and one dependent lookup (the words offered in question 5 depend on the component chosen in question 3). It must refuse to submit with a blank field, must be tested for offline behavior on site (fallback is the Monday mobile app, never paper), and writes to a Condition Log board that is re-created from a template every calendar year to stay clear of the 10,000-row cap. "Re-look" rows reference the prior year's row across boards, and a report card is generated by reading this year's board against last year's.

7. Email in two tools with a test you will run. SuperMail handles one-building / one-event / stamped-on-the-row email (proposals, document requests, the deadline notices, delivery emails, suspension and exit notices). monday Campaigns is the intended tool for the monthly issue IF it passes a six-point merge test you will run on the sandbox (per-recipient merge of a long-text account field; the good-standing-AND-Checked gate; one email per building to a multi-address recipient list; sent-stamp write-back; bounce visibility; holds at 100+ buildings). If it fails any point, the monthly issue also goes through SuperMail using formula-concatenated recipient columns.

8. It has to live inside a working business. Everything goes in a new workspace in the same account, reusing our existing Customer Board / CRM Accounts (which carry the QuickBooks customer IDs) and Contacts, conforming to the monday CRM structure, without disturbing roughly 1,500 live items and 130 automations on the existing boards.

9. An AI drafting layer reads and writes specific columns. Scheduled jobs (our side) read Monday and drop first drafts onto designated subitems and control-row fields. You do not build the AI, but you design the columns it reads and writes, confirm the API access, and make sure no alarm anywhere depends on it.

4. THE ARCHITECTURE (ALREADY DESIGNED — YOU BUILD IT, YOU DO NOT REDESIGN IT)

Workspace. A separate SHIELD workspace in our existing Monday account. Files live in Google Drive (the building's row carries its folder link); every sealed PDF is also attached to the subitem that released it.

Board A — Members. One item per building for the life of the membership and after it. Groups = lifecycle stages. Column families: sales (contacts, roles, PM firm, lead source, proposal button + file + DocuSign status, signed agreement); contacts (one required primary notice contact, up to three secondaries, an acknowledgement date, a formula → text recipients column, a bounce flag); membership (effective date, membership year, budget-adoption date mirrored from Board B, payment election, payment status from QuickBooks, Paid Months, a locked price field editable only by the owner after signing, vesting math, good-standing flag, a "SIRS on file?" gate, an Exit Date formula, Suspend? and Terminate? buttons restricted to the owner, account/sub-customer link, Drive folder link, file columns); mirrors (next deadline from B, open findings count from C, walk status from the annual-walk subitem); and the "⚡ Something happened" button.

Board A subitems — the deliverables table. Every event in a building's membership life is one subitem, typed (Deliverable · Monthly Update · On-call Log · Decision/Correspondence · Meeting), with a Trigger column (Clock = a date created it; Event = something happened), membership year, due date, owner, status, file/note, and sent stamp. Subitems accumulate on purpose — the list is the building's history. On exit, subitems due after the Exit Date are cancelled, never deleted.

Board B — Compliance Calendar. One item per deadline per building, connected to A. Deadline type; a Statutory? flag (Yes = guaranteed notices; No = tracked only); deadline date with two blank-states ("Date needed" at onboarding, "Confirm date" after an anniversary rollover); source document; Notice −12 / −6 / −3 formula dates with sent stamps; walk-window and walk-backstop dates on the budget-adoption row; owner; handled; escalation. Automations: starter rows spawn at GO; blank-date pings at +10 and +20 business days; anniversary rollover; notice send + stamp + 2-business-day escalation; 30-day "not handled" alert; notices stop only after the Exit Date. Notices never check good standing. This board is where the guarantee lives, so it is built to the highest standard of dependability in the system.

Board C — Condition Log [YEAR]. One board per calendar year; one item per observation, all buildings, connected to A. Fed by the inspector's phone form. Row types include defect, "no deficiency observed," from a prior report, reported-not-verified, project done without us (created by the event form), re-look (new row pointing at the old one), special walk, and outside-scope hazard (no grade). Closeout fields live on the same row. We supply the complete form vocabulary as an exhibit.

Board D — Monthly Issue. One item per month. Two shared text sections; the owner's approval status for the law item; the general manager's approval status for the season item; draft-ready / checked / sent / held / skipped counts; send stamp. It holds no building facts.

Forms (three). The intake form (Board A; creates the row, the contacts, and sets the "SIRS on file?" gate). The "⚡ Something happened" event form (event type · date · who told us · what happened · documents; ten event types, each routing to different created objects — including the two budget events and the membership-notice event that computes the Exit Date). The inspector's phone form (Board C, described above).

Integrations. DocuGen (proposal, scope-baseline exhibit, notice letters, suspension/termination notices, handoff letter shell) · SuperMail (templates and recipient logic as above) · DocuSign (envelope status → proposal status; scan-and-upload path also supported) · Webhooks → Make → QuickBooks Online (all invoice types, credit memo, payment and invoice-number write-back, Paid Months feed) · Google Drive (folder creation and link on the row, via Make) · monday Campaigns (subject to the merge test) · the AI drafting layer (scheduled jobs via the Monday API; columns designed by you, jobs built by us).

What is deliberately NOT built: separate sales or financial boards; per-building calendars; any meeting-format logic; any archive automation that moves Board C rows on a date.

5. SCOPE OF WORK — WHAT YOU DELIVER

Phase 0 — Specification review and gap analysis (hourly, capped). Read the builder pack. Walk the acceptance-test script on paper. Deliver a written gap list: every item in the spec that Monday cannot do natively, with your proposed approach for each (native automation vs. monday Workflows vs. Make) and the trade-off. Confirm plan limits that matter (automation actions per month, storage, Campaigns contact and send limits, subitem automation limitations, API rate limits). Confirm the Customer Board → CRM Accounts mapping. Deliver fixed-price quotes for Phases 1–6.

Phase 1 — Sandbox structure. The workspace; Boards A, B, C (as a yearly template) and D; every column with type, labels, and a description written as an operator instruction; groups; the subitem structure; connect-boards and mirror columns; every formula (Paid Months, vesting credit, Exit Date, notice dates, season offsets, recipient concatenation); the December board-rollover procedure written as a checklist. You build the entire system from the specification — nothing is pre-built or inherited. Our side independently audits each phase against the specification, column by column and automation by automation, before that phase is accepted; this is in addition to your own QA, not a substitute for it.

Phase 2 — Forms. The intake form; the "⚡ Something happened" form with its routing; the inspector's phone form built to our vocabulary exhibit, including the dependent lookup and the no-blank rule; an on-site offline test of the phone form with a written result.

Phase 3 — Automations. The GO spawn; the calendar and notice engine with stamps and escalations; blank-date pings; the monthly-issue hard hold and write-back counts; budget-season offsets and the "SIRS on file?" gate; the anniversary spawn and "Confirm date" rollover; the delinquency → suspension → termination → exit path with the two owner-only buttons; the exit cancellations and handoff spawn; same-row rejoin. Every automation documented in a register as one trigger → action sentence with its board and its purpose.

Phase 4 — Integrations. DocuGen templates; SuperMail templates and recipient logic; the DocuSign connection; the Make → QuickBooks Online scenarios with error handling, retries, logging, and write-back; Google Drive folder creation; the Campaigns merge test with a written pass/fail against the six criteria and the SuperMail fallback wired either way; the AI-layer columns and API access confirmed and documented.

Phase 5 — Acceptance test. One fictitious building run from intake through Year 3, including: one late document turnover, one blank-date escalation, one full monthly cycle with a held building, one Version-B budget season (proposed budget → alignment letter → adoption → state-filing row spawned), one anniversary rollover, one delinquency → suspension → cure, one exit with handoff, and one rejoin — with expected vs. actual at every step, screenshots or logs, and fixes until it passes. A second person on our side re-runs the script cold.

Phase 6 — Go-live and handover. Move the boards into the CRM structure; documentation (board dictionary, automation register, Make scenario docs with diagrams, a runbook for the monthly cycle and the December rollover); two recorded training sessions (one for our operations manager on input duties, one for our general manager on monthly production and release); 30 days of post-launch support with a defined response time.

6. WHAT WE PROVIDE

The builder pack: the architecture document, a plain-language step-by-step process narrative, an eight-part delivery narrative, the Build Spec (column by column, automation by automation, with the inspector-form vocabulary as an exhibit), and five maintained diagrams (boards explained, storyboard, service calendar, budget-season Gantt, infographic set). Admin access to a sandbox workspace and to our Make account. Read access to the existing boards as reference patterns — our two newest boards are the style to follow. One decision-maker (the firm's president) who answers questions the same day. A 30-minute weekly check-in. Test data for the fictitious building.

Disclosure is staged. Shortlisted candidates receive a plain-language process narrative and two system diagrams under NDA, which is enough to quote Phase 0 and give a preliminary range for the rest. The full builder pack goes to the selected builder at the start of Phase 0.

7. OUT OF SCOPE

Writing the AI drafts themselves (we build that layer); the Work Management → CRM migration of our existing boards; contract or marketing content; QuickBooks bookkeeping; purchasing Monday or Make licenses (we buy; you advise on what is needed).

8. REQUIRED SKILLS AND EXPERIENCE

• Monday.com, advanced: formulas; subitems and automations that act on subitems; connect-boards and mirror columns; date-arrives automations with offsets; buttons with owner-only restrictions; forms with conditional logic; the monday CRM product structure; monday Campaigns; dashboards. Working knowledge of monday Workflows and the GraphQL API.
• Make (make.com), advanced: webhook-triggered scenarios; routers; error handlers and retries; data stores; the Monday and QuickBooks Online modules; write-back patterns; scenario documentation.
• DocuGen, SuperMail, and the DocuSign integration for Monday — production experience, not just familiarity.
• QuickBooks Online: invoices, credit memos, customers and sub-customers, payment status.
• Preferred: Google Drive via Make; experience with deadline-driven or compliance-driven service businesses; prior testing of monday Campaigns per-recipient merges; contract-quality written English (column descriptions and documentation are read by our staff every day).

9. WORKING RULES (NON-NEGOTIABLE)

1. NDA and IP assignment before you see the builder pack. Every board, automation, form, template, scenario, and document belongs to GBE&C and is built under GBE&C logins and GBE&C's Make account. No personal accounts, ever.
2. Priorities, in order: dependable, then accurate, then efficient. When you must choose, choose the boring, reliable way.
3. Never fake it. If Monday cannot do something the spec asks for, say so in writing and propose the alternative. A silent workaround that fails in month eleven is the worst outcome this project can have.
4. No alarm depends on AI or an outside scheduler. Every guaranteed notice is a native Monday automation.
5. Nothing client-facing sends without a human status upstream.
6. Document as you build. The register and the dictionary are deliverables, not afterthoughts.
7. The build never blocks the business. We may sign real members before you finish; we will run them by hand and migrate rows in. Plan for that migration.
8. Confidentiality. The program design is proprietary and unreleased.

10. ENGAGEMENT STRUCTURE

Phase 0 is hourly with a cap. Phases 1–6 are fixed price per phase, quoted at the end of Phase 0, paid on acceptance of each phase. Post-launch support after the included 30 days is an hourly retainer. Target timeline: Phase 0 starts on hire; we are aiming for the acceptance test to pass within roughly 12 weeks of Phase 0 sign-off, with the final schedule set with the selected builder at the end of Phase 0. At least four hours of daily overlap with U.S. Eastern time. Communication in English. Contract and payment through Upwork.

11. HOW TO APPLY

Start your proposal with the word STAMPED so we know you read this far. Then answer, briefly and specifically:

1. Describe the largest Monday.com system you have personally built — boards, approximate automation count, integrations, what broke after go-live, and how you fixed it.
2. Describe a Make → QuickBooks Online scenario you built that wrote payment status back into Monday. What failed, and how did you handle errors?
3. How would you implement this: "Twelve months before a date on a connected calendar row, send a templated email to a contact stored on the parent building row, stamp the send date and time on the calendar row, and alert two named people if the stamp is still blank two business days later." Say what is native, what needs Make, and where the traps are.
4. Have you tested monday Campaigns merging per-recipient fields from CRM accounts? What happened?
5. Your availability, your proposed Phase 0 approach, and your Phase 0 hourly rate and cap.

Redacted screenshots or short recordings of prior builds are welcome. Proposals that do not answer the five questions will not be reviewed.

Virtual Assistant — Delilah Agency (Remote, Part-Time)

Base hourly pay + commission

Delilah Agency is a social media repost service, with connected ventures in music marketing (OTL Music) and a recording studio. We’re looking for a self-starter to own outreach and act as a right hand across the business.

What you’ll actually be doing:

• Research — Finding qualified leads across our services: local businesses/brands for Delilah’s Instagram growth packages, artists/labels/publishers for OTL Music’s playlisting, audiobook, and podcast production services, and potential studio clients or talent opportunities. Building and maintaining organized contact lists.
• Outreach — Writing and sending cold emails and Instagram DMs to those leads: pitching our services, following up on past conversations, and reaching out for partnership or booking opportunities. This includes both introducing new prospects and re-engaging warm ones.
• Reporting — Regularly reporting back to me on who you contacted, what responses came in, and what needs my attention or a decision. I want visibility without having to ask for it.
• General task execution — Whatever else comes up day to day: scheduling, light admin, research requests, updating trackers/CRM, helping coordinate between the different ventures. You take a task and run with it.

Who we’re looking for:

• A genuine self-starter — you don’t need step-by-step instructions to get moving, and you flag questions instead of stalling
• Experience in music marketing, business/brand marketing, or talent management — you understand how to identify opportunities and speak the language of artists, labels, or brands
• Strong, natural writing — outreach that reads like a person, not a template
• Comfortable owning outbound outreach as your core responsibility, not just responding to inbound
• Organized enough to manage multiple lead lists and follow-up sequences without things falling through the cracks

Delilah Agency is a growing local Instagram growth/repost service (with connections to OTL Music and a recording studio) looking for a sharp, self-directed VA to help run outreach and day-to-day ops.

Perfect for individuals familiar with music marketing and/or business marketing and/or talent management/marketing

What you’ll do:

• Research leads (local businesses, artists, potential reposters/partners) across our services
• Send outreach emails and Instagram DMs — cold outreach, follow-ups, partnership inquiries
• Handle ad-hoc tasks as assigned — scheduling, research, light admin, CRM/spreadsheet updates
• Keep outreach organized and report back on responses/results

You’re a good fit if you:

• Write clean, professional, non-robotic emails and DMs
• Are comfortable doing outbound outreach (not just answering inbound)
• Can self-manage a task list with minimal hand-holding
• Are organized and communicate proactively — no need to chase you for updates
• Bonus: familiarity with Instagram growth/marketing, Airtable, or Make.com

Details:

• Part-time, ~10–20 hrs/week
• Pay: hourly + commission on outreach that converts
• Remote — work from anywhere
• Report directly to the founder

I'm looking for an experienced AI Voice Agent developer to build an AI receptionist that can handle inbound business calls and provide a natural, human-like conversation experience.
The agent should be able to answer calls, understand caller intent, answer questions, collect information, qualify leads, schedule appointments, transfer calls when needed, and handle common customer inquiries.
Key Requirements
• Build and configure an AI voice receptionist
• Natural, low-latency voice conversations
• Inbound/outbound calling capabilities
• Call qualification and information collection
• Appointment scheduling and calendar integration
• Call transfer/escalation to human agents
• Integration with CRM and other business systems
• Knowledge base/RAG integration where required
• Call logs, transcripts and basic analytics
• Reliable handling of interruptions, silence, unclear responses and unexpected questions
Preferred Technologies
Experience with platforms such as Twilio, Vapi, Retell AI, ElevenLabs, OpenAI, Claude, n8n, Make, APIs and webhooks is highly preferred.
I'm looking for someone who can design the workflow, recommend the right technology stack, build the solution, test it thoroughly, and provide clear documentation.
This can potentially turn into a long-term engagement for improving and expanding the voice AI system.
Please include examples of similar AI voice agents or AI receptionists you've built when applying.
Please include these keywords/skills/tags
• AI Voice Agent
• AI Receptionist
• Voice AI
• Conversational AI
• AI Chatbot
• OpenAI
• Twilio
• Vapi
• Retell AI
• ElevenLabs
• n8n
• Make.com
• API Integration
• CRM Integration
• Appointment Scheduling

About the Project

We are looking for an experienced ActiveCampaign expert to help us set up and configure our ActiveCampaign environment for the first time. Close collaboration is important to us: we want to build the automations together so that we can independently extend, adjust, and manage them in the future. We are therefore not just looking for someone to implement the work, but also an advisor who can communicate their expertise clearly.

Tasks

Joint setup and configuration of ActiveCampaign
Developing a sensible, scalable structure (lists, tags, segments, fields, and automations)

Building automated email sequences and follow-up processes
Setting up automated feedback and review forms

Creating reminder and escalation loops for cases where forms or feedback are not completed

Setting up triggers and conditions for different processes
Advising on best practices and a future-proof system structure

Supporting documentation of key processes

What Matters to Us

Strong expertise in ActiveCampaign

Experience with marketing and process automation

Ability to explain technical concepts in an understandable way
Willingness to carry out the setup together with our team and answer questions directly

Structured, solution-oriented way of working

Goal of the Collaboration

By the end of the project, we want to have a properly configured ActiveCampaign environment and be able to independently create and adjust future automations. We therefore place great importance on knowledge transfer and close collaboration during implementation.

Please share with us:

A brief introduction
Your experience with ActiveCampaign
References or comparable projects
Your hourly or fixed budget
Your time availability