AI App Development Projects on Upwork

Website: Thrice.Agency

We're an AI-first marketing company with several SaaSes and many clients.

We need an AI-first developer to keep up with demand. We have a US-based dev who's been with us for years who you will work with, as well as regular reporting and training with our founder who has the most technical acumen on staff.

Small organization, 4 heads atm. Rate is an intro rate during probation for at most 6 weeks. Time fraud is immeniently detectable with immediate firing. Flexible hours, but must have at least 4 hours per day overlapping with 11a-6p MT.

Cursor is the IDE, little flexibility there, have to use it. We'll comp the AI costs. Development serves business objectives, so one must be comfortable working in the timelines and budgets of our SMB clients that consider perfect the enemy of the good.

You don't need to be perfect or even talented. You need an analytical and structured mind that's able to see multiple solutions and choose the most cost-effective route based on cost-benefit analysis.

You don't need to be a pro at coding or Cursor. There's little precedent for the new wave of AI coding that is overtaking all other forms. We're extraordinarily good and will be imparting that knowledge to you, arguably better than any other training for this new wave of development. It's essential that you're trainable and don't flaunt expertise where you have none.

Daily standup at 10:30A MT in Gather.town.

We're flexible, fun, and open-minded. There's chatter at our digital workspace. This is understood as part of the job by the preexisting team, design to promote unity and morale. It's still essential to get work done, on time, and ask for help when needed.

The org is small and less structured than large business. This is fun and a great fit for someone who values personal connection and teamwork. It's a horrible fit for someone looking to coast and collect a paycheck, mainly because they'll be booted quickly.

Currently maintaining quarterly raises. Org is doing well and will likely be able to triple this intro rate over the coming year, incrementally as we scale. That, of course, depends on your performance since we'll be counting on you.

Gigabite.ai is building global supply chain infrastructure for food distributors. www.gigabite.ai. We connect U.S. businesses with global suppliers using AI-powered matching and real-time freight/tariff calculations.

Small team, moving fast. You'll work directly with founder, 3rd-time, operate a $100M import company and our architect (owns core algorithms).

The Role
Architect handles quotation logic, freight/tariff, data pipelines. You own everything else:

Frontend: Distributor-facing experience — catalog, savings reports, quote flows. JavaScript, React/Vue. Heavy API integration.

Backend: Third-party integrations, internal tools, data import/export. Python. Work from existing patterns.

Quality: Enterprise customers find every broken flow. You catch problems first.

You see the full picture, make decisions, ship.

The ideal candidate has built data-heavy B2B tools — not marketing sites, Fluent in frontend (JavaScript, React/Vue, state management) and backend (Python, APIs)

Read existing code, understand patterns, extend without hand-holding

Detail-oriented. Notice when something's off before users complain.

Treat AI as a multiplier, not a crutch — you use Claude to move faster, but you own the output
Ask good questions, then run.

Plus if: Enterprise software experience. Logistics, supply chain, or distributor background.

I’m looking to build a complete, fully branded Greenline feeding system for Farming Simulator 25 and need an experienced modder who can take it from concept to fully functional in-game machinery. The pack will include Wasserbauer livestock robot feeders and batching units, the full Cormall hay-processing line, and the Rovibec Ranger feed-pusher robot, all skinned and branded to match Greenline’s identity.

Scope and Priorities
• Wasserbauer robots – Autonomous navigation is the top priority. Once reliable path-finding is achieved, they should handle feed dispensing, batching functions, and standard maintenance cycles.
• Cormall equipment – We begin with the conveyor system. It must support automatic start/stop triggers and integrate smoothly with other Greenline units. The mill and remaining components follow afterward.
• Rovibec Ranger – Should patrol feed alleys, push leftover ration back into the manger, and follow existing cow-AI schedules.

Deliverables

High-quality 3D models with proper LODs, dirt/wear maps, and full Greenline branding.

XML and LUA scripts that integrate with FS25 systems including feeding, animal growth, and economy.

Fully tested autonomous movement for the Wasserbauer robots.

Cormall conveyor with automatic start/stop logic and accurate machine-to-machine handoff.

A complete ModHub-ready package with icons, store data, and localisation.

Acceptance Criteria
The project is complete when all machines appear in the in-game store, connect and operate without errors, and can run a full unattended feeding cycle for at least three in-game days.

Important
Only apply if you have worked on similar Farming Simulator projects and can provide links to live mods, live projects, and 3D model samples.
Preference will be given to applicants who can share relevant examples upfront.

I am looking for a developer to create a simple AI-powered text summarization script in Python. The script should accept a text input (pasted by the user), send it to an AI API (OpenAI or similar), and return a clear and concise summary of about 3–5 sentences.
Project Requirements:
Build a clean, easy-to-run Python script
Implement an API call to an AI model for text summarization
Allow the user to paste any text as input
Output a short, consistent summary (3–5 sentences)
Provide brief instructions on how to install dependencies and run the script
Code should be simple, readable, and ready for testing
Deliverables:
Python script (.py file)
Simple instruction manual (README or text)
Optional: GitHub link for the final code

Boost Health Insurance seeks a mid-level Voice AI Optimization Specialist to maintain and enhance our live ElevenLabs-powered voice AI system that contacts existing customers for policy renewals and follow-ups, as well as recovers incomplete leads for sales conversion.

This is a full-time, independent contractor position reporting to our Director of Operations. You'll be responsible for continuously improving our outbound dialing campaigns with a focus on measurable sales outcomes.


Primary Responsibilities (70% Optimization):
- Monitor and optimize conversion rates across customer follow-up and lead recovery campaigns
- Analyze call performance metrics and identify opportunities for improvement
- Refine conversation scripts, objection handling, and call flows based on data
- Reduce response times and improve customer experience
A/B test different approaches to maximize engagement and transfers to sales agents
- Track KPIs including connection rates, conversation completion, transfer rates, and conversion metrics
- Secondary Responsibilities (30% Development):
- Build new voice AI capabilities and conversation paths
- Manage ElevenLabs API configurations and optimizations
- Oversee SIP integration between dialer and voice AI bot
- Implement new features based on campaign needs and performance insights
- Integrate with CRM systems as needed


What Success Looks Like:
- Measurable improvements in conversion rates within first 90 days
- Faster, more natural call response times
- Better objection handling leading to higher transfer/completion rates
- Scalable, repeatable processes for ongoing optimization


Required Skills & Experience:
Must Have:
- Proven experience optimizing outbound dialing campaigns for sales with measurable results
- Strong understanding of sales metrics, conversion funnels, and campaign performance analysis
- Experience with voice AI platforms and conversational AI optimization
- Hands-on experience with API management and SIP/telephony integrations
- Data-driven approach to testing, iteration, and optimization
- Ability to work independently with minimal supervision
- 2-4 years relevant experience in voice AI, sales automation, or conversational AI


Nice to Have (Bonus Points):
- Experience with ElevenLabs voice AI platform
- Background in insurance or financial services industry
- AI prompt engineering and LLM optimization expertise
- Experience with CRM integrations (Salesforce, HubSpot, etc.)
- Familiarity with outbound dialer platforms
- Track record of improving sales conversion rates in automated campaigns


Scope of Work:
Work Arrangement:
- Full-time contractor position (40 hours/week, ~160 hours/month)
- Remote/flexible location
- Reports to Director of Operations

Budget:
$35-55/hour depending on experience
Approximately $7,200-8,800/month


Infrastructure:
- You will manage: ElevenLabs API, SIP configurations, conversation design
- You will NOT manage: Server infrastructure (handled by IT team)


Success Metrics:
- Conversion rate improvements
- Call response time reduction
- Objection handling effectiveness
- Customer satisfaction and transfer rates
- Campaign ROI and cost-per-conversion


To Apply: Please include:
- Examples of outbound voice/dialing campaigns you've optimized with before/after metrics
- Your approach to improving conversion rates in automated voice systems
- Relevant experience with voice AI platforms and telephony integrations

Create a clear AI/LLM system design PDF for a production-ready chat app with RAG, guardrails, and monitoring. This is documentation-only and must be practical enough for engineers to implement.

Requirements to cover in the PDF:

High-level architecture + data flows (ingestion → indexing → chat)

Components (API, orchestration, vector DB, storage, cache, auth/RBAC)

Cost controls (caching, model routing, token limits)

Security/guardrails (prompt injection + PII handling)

Deployment outline (AWS preferred) + key KPIs/monitoring

Failure modes + mitigations

To apply (answer in proposal):

5 key components you’d include

3 top security risks in LLM apps

Confirm you can deliver a PDF within 48 hours

We are seeking a skilled freelancer to build a replica of a custom site using Base44 and/or Lovable. The ideal candidate must have experience in Base44 and/or Lovable, AI native apps, AWS, SEO optimization, and ensuring site security. The site must be responsive and fully functional.

Must be okay with the budget

• Describe your recent experience with similar projects
• Please list any certifications related to this project

Project: Golf equipment analysis app using Claude API for club identification and grading
Current Stack:

FlutterFlow (frontend)
Firebase (Firestore, Cloud Functions, Auth)
Claude API (Anthropic) for AI features
Node.js Cloud Functions (can also use Python)

Situation:
We have a working app with existing Cloud Functions for club identification and bag grading. Our FlutterFlow developer needs guidance on AI best practices - specifically prompt engineering, API response handling, and data flow optimization. We're experiencing:

Claude API sometimes hallucinating club specifications
Data not consistently flowing from frontend to Cloud Functions
Need to optimize prompts for more reliable identification

What We Need:
Review our existing code and provide actionable recommendations. May also implement fixes directly (scope TBD).
Deliverables:

Review existing Cloud Functions and Claude API implementation
Written recommendations document
Available for Q&A with our FlutterFlow developer
May implement code changes directly (scope to be determined)
Join shared Slack channel for team communication

Required Skills:

Claude API or similar LLM APIs (OpenAI, etc.)
Firebase Cloud Functions (Node.js or Python)
Prompt engineering
Debugging data flow issues

To Apply:

Relevant LLM API experience
Example of a prompt engineering challenge you solved
Availability over next 2-3 weeks
Hourly rate

Salary: Up to $12,000/year ($1,000/month) for full-time work. Monthly payment.

We’re hiring a full-stack developer to join us on our AI chat-to-app builder e.g. Lovable, Base44, etc. Product is already profitable and on a great trajectory (will show you).

• You’ll work side-by-side (async) with a team that’s built, sold, and acquired SaaS companies before (multiple times 🙂). We know how to scale. Now we need someone who can out-ship everyone else

• We ship fast, chat async, and care only about results and users - not processes, very few meetings.

• Need you to be a beast, expect to work on weekends too. It's mandatory, since the whole team does that too - we move quickly and ship constantly. Need you to be hungry. Don’t apply otherwise

• I need an output-driven person, one who is more concerned about what makes users say “This is what I need! This solves my problem!” And less about what tech we use, losing themselves in the minutiae details
• We’ve reached $4.7k MRR after a couple of months. If no one would’ve cancelled their subscriptions, we would’ve been at $11k MRR… and ofc even more due to word of mouth. This is pure validation that our marketing (myself and my cofounder) works. The bottleneck isn’t demand, it’s the foundation of the product. Once that clicks, our marketing floodgates will amplify a great product. NOTE: There are few other products on the market that are more guaranteed that they’re gonna “work”… IOW: There are few products on the market that can guarantee you more that, once your delivery works, it will mean real growth. We have the marketing chops, and we can do even more once product confirms it.

Our stack:
• Next.js
• Vercel AI SDK v5 (used heavily)
• Neon database (main app)
• Turso database (sandboxes)
• ShadCN UI
• Modal.com

Optional: Vite, Railway, Prisma, Turso, Langchain, Voltagent, Turborepo, Express

You’ll be handling both frontend and backend, with plenty of room to own features end-to-end. When applying, start the application with "Not copy-pasted" so I know you actually read this.

We’re looking for someone who:
• Is output-driven and obsessed with user impact
• Doesn’t need hand-holding - just context and ownership
• Can handle real-world pressure, fast iterations, and async work
• Works weekends, late nights, whatever it takes - because they love the craft
• Wants to build something big, not just “have a job”

To apply:

1. Shortly: What’s your experience with the mentioned stack (all items)?

2. Shortly: Do you know Lovable/Base44? Have you used them?

3. Link your Github pls

4. BONUS: Got any live application URLs or side/personal projects you’ve built? Send them pls - this will make you stand out

We’re looking for a highly technical developer or small expert team to build the first version of an AI-driven SaaS Knowledge OS — a platform that ingests business documents, emails, and structured ERP data, and transforms them into a unified, queryable knowledge graph.

This project is not a simple CRUD SaaS. You will be designing intelligent data integration, entity resolution, and graph-aware enrichment workflows that allow an LLM to understand and connect business information across multiple systems.

This sprint will be a two-week MVP aimed at proving the core concept:
LLM-powered entity extraction, enrichment, and deduplication into a Neo4j-based knowledge graph.

White-Label, AI-First Health Ecosystem
Company: Shae Group
Engagement Type: Agency / Delivery Partner
Location: Remote (Global)
Engagement Start: ASAP
RFP Due: by 10am ACST 19th Decemeber. All interviews will be conducted on Friday 19th December between 10am-9pm ACST (Adelaide, Australia)

About Us

We are building a next-generation, AI-first, white-label health intelligence platform designed to support consumer wellness, clinical care, enterprise wellbeing, and long-term global research.

At the center of the ecosystem is a Conversational AI and Analytics Dashboard that delivers personalised, proactive guidance across health, wellness, lifestyle, performance, and behaviour. To power this intelligence, we are building modular, reusable infrastructure components that can be deployed as:
- SaaS (white-label apps)
- API licensing
- Customer-hosted IaaS / IaC deployments

We are now seeking an experienced agency partner to design and build our Wearables & Consumer Health Data Module.

The Opportunity

You will design and deliver a production-grade Wearables Module that:

- Ingests consumer wearable and health data
- Normalizes it into a canonical health data model
- Exposes clean APIs for downstream AI, analytics, and applications
- Is fully white-label, multi-tenant, and reusable across all clients

This module will become a core building block of the Shae ecosystem and will be reused across multiple products and customers.

Scope of Work (Phase 1)
1. Wearable & Consumer Health Data Ingestion
Design and implement secure ingestion pipelines for consumer health data, including (but not limited to):

- Apple HealthKit
- Google Health Connect

Key requirements:

- OAuth-based consent and permissions
- User-centric consent revocation
- Secure data ingestion (polling and/or webhook-based where available)
- Idempotent, fault-tolerant ingestion

2. Supported Data Categories

The system must be capable of supporting a broad range of consumer wearable signals, including:

- Activity & movement (steps, distance, activity types)
- Sleep (duration, stages, efficiency)
- Heart metrics (heart rate, resting HR, HRV)
- Respiratory metrics
- Energy, exertion, recovery proxies
- Stress indicators (where available)
- Body metrics (weight, body composition where supported)
- Raw vs derived metrics distinction
- Time-series data at multiple granularities (minute, hourly, daily)

A detailed wearable data taxonomy will be provided during the PRD phase.

3. Data Normalisation & Canonical Models

- Design a canonical internal data model that abstracts device-specific schemas

- Handle:
-Time alignment
-Unit normalization
-Data completeness and confidence
-Duplicate and partial records

- Support historical backfill and incremental updates

4. API & Integration Layer
Deliver clean, versioned APIs that allow other systems to consume wearable data, including:

- Digital Twin / AI orchestration
- Surveys & assessments
- Dashboards and scorecards
- Product recommendation engines
- (Optionally) clinical systems via future EMR integrations

APIs should support:

- Time-range queries
- Aggregations
- Per-tenant configuration
- High read throughput

5. Multi-Tenant & White-Label Architecture

The module must support:

-Strict tenant isolation
- Per-client feature toggles
- Client-specific wearable enablement
- Deployment via:
- Managed SaaS
- API licensing
- IaaS / Infrastructure-as-Code

6. Security, Privacy & Compliance

- Encryption at rest and in transit
- Audit logs for data access
- Clear consent lifecycle management
- Support for global privacy requirements (GDPR-aligned design)
- Secure data deletion and export flows

Explicitly Out of Scope
To avoid ambiguity, the following are not part of this engagement:

- AI recommendation logic
- Scoring algorithms
- Survey logic
- Front-end mobile or web UI
- EMR ingestion (clinical systems)
- Business intelligence dashboards

These are handled by separate teams.

What We’re Looking For

Required Experience

We are looking for agencies with proven experience in:

- Consumer wearable and health data platforms
- Apple HealthKit and Google Health Connect integrations
- Multi-tenant SaaS architectures
- API licensing (PaaS) & on prem (IaaS) coding/deployment experience (or at minimum, have awareness of these, so plans can include these)
- High-volume time-series data systems
- Secure, privacy-first data pipelines
- API-first backend development

Strongly Preferred
- Experience in health, wellness, or regulated data environments
- Familiarity with standards such as FHIR (even if not used directly)
- Experience building reusable, white-label platform components
- Cloud-native, production-grade delivery (not MVP-only work)

What to Include in Your Proposal

Please include the following in your submission:

Your proposed approach
- Architecture overview
- Key design decisions
- How you future-proof the module

Technology stack

- Languages, frameworks, databases
- Managed vs open-source components
- Trade-offs and alternatives

Team & resourcing

- Roles and number of people
- Seniority
- Ability to add parallel resources

Timeline & milestones

- High-level delivery plan
- Milestone breakdown
- Impact of increased resourcing on timeline and cost

Costing

- Total estimated cost
- Cost per milestone
- Assumptions and exclusions

Relevant experience

- Examples of similar systems built
- Case studies or references

Engagement Details

Proposal submission window: 5 business days. All interviews will be conducted on Friday 19th December between 10am-9pm ACST (Adelaide, Australia)

Shortlisting: Selected agencies will be invited to a technical deep-dive

Next phase: Detailed PRD and architecture alignment with shortlisted partners


Why This Is Interesting

You will help shape a foundational platform component, not a one-off integration

Your work will be reused across multiple white-label products and clients

You’ll be working alongside a deeply thought-through AI-first product vision

Long-term partnership opportunities beyond Phase 1

Location: Remote / async-friendly. Do you live in GMT+5-GMT+7 = do not apply. We live in opposite timezones to you, so it’s too tricky to book professional calls!

About Mirage:
Mirage is a new kind of interactive mobile experience where music, AI, and creativity collide. Built by a team of music lovers and tech innovators, we’re developing an app that empowers users to remix, create, and share music in entirely new ways — using next-gen tools and stunning interactive design.
Launching soon, Mirage blends cutting-edge AI with intuitive UX and a playful, gamified layer to engage the next generation of creators and fans.

We’re Hiring: AI Engineer / Data Scientist (Audio):
We’re looking for a talented AI Engineer / Data Scientist specialized in Audio to join our core tech team. You’ll help us push the boundaries of what’s possible in AI-generated music, voice synthesis, and audio signal processing, powering the unique gameplay experiences of Mirage.

Your Mission:
Develop and optimize deep learning models for audio/music generation (tracks, voices, instruments)
Work on voice cloning, source separation, and real-time audio processing
Collaborate with designers & developers to bring AI features seamlessly into gameplay
Research and experiment with state-of-the-art audio ML architectures (transformers, diffusion models, GANs)
Ensure scalable deployment of AI models into mobile gaming environments.

Your Toolkit:
Strong background in Machine Learning / Deep Learning, especially for audio
Proficiency in Python, PyTorch/TensorFlow, and audio libraries (Librosa, Torchaudio, etc.)
Knowledge of speech synthesis, music information retrieval (MIR), and DSP
Experience with LLMs for multimodal audio+text tasks is a plus
Familiarity with cloud deployment (AWS, GCP, Azure) and MLOps pipelines

What We Offer:
Work on a game-changing music & gaming AI startup with global ambitions
A unique chance to combine cutting-edge AI research with massive entertainment impact
Collaborate with top music, gaming, and AI innovators worldwide
Flexible, remote-friendly work with an international, passionate team
If you’re passionate about music + AI and want to build technology that millions of players will experience, Mirage is your stage.

We are seeking a skilled developer to create a web application that captures streaming video input, such as from a camera. The application should include functionality to detect individuals in the video feed and overlay virtual accessories, like glasses or other items. The ideal candidate will have experience with video processing, computer vision, and real-time application development. If you're passionate about creating engaging applications and have a strong understanding of the necessary technologies, we want to hear from you!

a perfect example of the solution would be that the user is taking an ipad or similar tablet and he/she sees herself with a (different) pair of glasses or earrings or any other improvement of her face straight away.

Please refer to the brief in the attached file. Important - I work a lot with LLMs and am immediately put off by responses written by them.

I am looking for someone who I can collaborate with and work with. This first project is small but would lead to partnering with me on our ongoing tech roadmap. Therefore please read the brief properly, and reach out with sensible questions.

We are looking for a developer to build an MVP of an Automated Niche Research Platform.

Project Overview
Users will enter any niche on the website, and an automated workflow (Make.com or n8n) will generate a structured market analysis.
Your task will be to set up the automation logic, connect AI models, and integrate everything with a simple frontend built on v0.dev (React/Next.js).

The platform will display dynamic sections such as:
- Demand and pain point analysis
- Niche decomposition
- Competitor overview
- Top-10 competitors
- Product and marketing analysis
- Recommendations for entering the niche

We already have the UX structure prepared. We need you to implement the system end-to-end.

Project timeline: 1 month.

Deliverables
- A functioning web app (Next.js + v0.dev) deployed on Vercel
- Automation workflow in Make.com or n8n
- Integration between frontend and automation via webhooks / API
- Database setup using Supabase, Neon
- Clean and understandable project structure
- Documentation for future extensions of the platform

Required Skills
- Experience with Make.com or n8n (webhooks, API calls, AI workflows)
- Strong React / Next.js skills (experience with v0.dev is a plus)
- Ability to handle API integrations and work with structured JSON
- Experience with Supabase / Neon and basic SQL
- Familiarity with Vercel deployment
- Understanding of AI integrations (OpenAI / Gemini / Claude)
- Python is optional bonus, not required

How We Work
- Fast and clear communication
- We provide all prompts, logic, and expected data structures
- We prefer iterative delivery (MVP → improvements)
- We give quick feedback to keep the project moving efficiently

If you have experience building automation-heavy AI tools or modern web apps, we’d love to hear from you.
Please share relevant examples and briefly describe how you would approach this project.

• Describe your recent experience with similar projects
• What frameworks have you worked with?

I am working on a small poc conversational chat bot
It has local vector db, English to promql query generator, tool calling like kubernetes api and other command line
Ollama qwen2.5 model used

I am still not perfect in my vibe coding the application but I want someone to clean it up because the response and results are not clean

My usecase is the agent has to query Prometheus metrics, kubernetes API and other tools to do root cause analysis

So it has to work locally
Multi tool calling and summarized with action points