Freelance python jobs for world-class Developers

We are looking for technical consultants that could help build ADT (admit, discharge, transfer) data pipelines into Snowflake, and then further into our care management platform called MHNConnect. We have recently built our Snowflake data platform but have upcoming projects with ADT vendors in CA & IN markets where we are looking to ingest/display their HL7 data for our patient population.

I wanted to understand feasibility, costs & implementation timelines for this request.

We are looking for an AI Engineer to help build and deliver AI-powered solutions for clients.

Projects include:

AI chatbots (customer support / internal tools)
Workflow automation
Integration with APIs and existing systems

Requirements:

Experience with OpenAI / GPT APIs
Experience building real applications (not just models)
Backend experience (Python or Node.js)
Ability to work on small to mid-size projects

Nice to have:

Experience with LangChain / RAG
Experience with automation tools (Zapier, Make)

We are looking for a long-term collaboration.

Job Title:
AI Developer Freelancer (Experimentation + AI-Native Product Implementation)

Job Description:
We are looking for a freelance AI Developer to help us experiment with AI models and implement AI-native product features.

Your role will involve testing different models, comparing performance, improving prompts and workflows, and turning working ideas into usable product features. This is a practical, hands-on role for someone who can move from experimentation to implementation.

Responsibilities:
- Experiment with different AI models, APIs, and workflows
- Evaluate model output quality, latency, and cost
- Improve prompts, system instructions, and agent behavior
- Build and implement AI-native product features
- Develop prototypes and production-ready backend logic
- Connect AI systems with APIs, databases, and external tools
- Help design practical use cases for LLMs and agentic workflows
- Debug issues and improve reliability of AI features
- Document findings, experiments, and implementation results

What We’re Looking For:
- Experience working with LLMs and AI APIs
- Strong skills in Python or TypeScript
- Experience building AI-powered applications or workflows
- Understanding of prompt engineering, evaluation, and iteration
- Ability to compare models based on quality, speed, and cost
- Experience integrating AI with backend systems, tools, or databases
- Comfortable working independently and testing ideas quickly
- Clear communication and practical problem-solving mindset

Nice to Have:
- Experience with agent workflows or multi-agent systems
- Experience with RAG, vector databases, or tool calling
- Experience with product prototyping and AI feature deployment
- Familiarity with OpenAI, Anthropic, Gemini, or open-source models

Scope of Work:
- Run experiments on models and workflows
- Recommend the best options for our use cases
- Implement selected AI-native features into our products
- Support iteration and improvement based on testing results

Engagement Type:
Freelance / Project-based

Proposal Requirements:
Please share:
- Relevant AI projects you have built
- Models/tools you have worked with
- Your preferred tech stack
- Examples of experimentation or AI product features you implemented

• Describe your recent experience with similar projects

This project is expected to be completed within 2 weeks from the start date.

1. Project Description
We are looking for a Senior Backend Developer to lead a critical refactoring and enhancement phase of our data-driven application. Our current architecture relies heavily on standard REST APIs, but as our real-time data needs have grown, we need to transition specific high-frequency data streams to WebSockets and MQTT.
The core of our stack is built on Python Flask, utilizing PostgreSQL for relational data and InfluxDB for time-series metrics. Everything is hosted on AWS. You will be responsible for ensuring our backend can handle persistent, bidirectional communication while maintaining a high standard of documentation and performance.

2. Problem List + Solution + Final Result Expectation
- Current Problem: High Latency/Polling: Our frontend/IoT clients currently poll REST endpoints for updates, causing unnecessary overhead and delayed data visualization.
Proposed Solution: Refactor high-frequency data routes to WebSockets and implement an MQTT broker (e.g., AWS IoT Core or EMQX) for device communication.
Final Result Expectation: A low-latency, real-time system where clients receive updates instantly via Pub/Sub or persistent sockets without constant polling.

- Current Problem: Outdated REST Logic: Some existing REST endpoints are inefficient, poorly structured, or lack proper validation.
Proposed Solution: Audit, modify, and optimize existing REST endpoints. Implement Marshmallow or Pydantic for better data validation/serialization.
Final Result Expectation: Clean, high-performance REST APIs that follow best practices and integrate seamlessly with the new real-time components.

- Current Problem: Documentation Gaps: Existing Swagger/OpenAPI documentation is incomplete or doesn't reflect recent changes, making it difficult for the frontend team.
Proposed Solution: Update and maintain comprehensive Swagger (OpenAPI) documentation for all REST endpoints and document the new WebSocket/MQTT event schemas.
Final Result Expectation: A fully interactive Swagger UI that serves as a "single source of truth" for the entire development team.

- Current Problem: Database Bottlenecks: Heavy read/writes to InfluxDB and Postgres are not fully optimized for the new real-time flow.
Proposed Solution: Optimize queries and ensure the backend efficiently handles the transition of data from MQTT/WebSockets into InfluxDB and PostgreSQL.
Final Result Expectation: Stable database performance with high data integrity even under increased real-time message loads.

3. Additional Job List
Infrastructure Support: Assist in configuring or fine-tuning AWS resources (EC2, RDS, or IoT Core) to support the new protocols.
Error Handling & Logging: Implement robust error handling and centralized logging to track socket disconnects and MQTT message delivery failures.
Unit & Integration Testing: Write comprehensive tests to ensure the refactored code doesn't break existing functionality (Pytest preferred).
Security: Ensure WebSocket connections and MQTT topics are secured via proper authentication (JWT/OAuth2) and TLS.

Required Technical Stack:
-Language: Python (Expert)
-Framework: Flask (including Flask-SocketIO or similar)
-Protocols: REST, WebSockets, MQTT
-Databases: PostgreSQL, InfluxDB
-Cloud: AWS (EC2, RDS, IoT Core)
-Docs: Swagger

• Describe your recent experience with similar projects
• Include a link to your GitHub profile and/or website
• What frameworks have you worked with?

# Senior AI Systems / Prompt Engineer

Remote · Format negotiable: full-time / part-time / project-based

## About Us

We are a fast-growing logistics company with a strong in-house IT and AI infrastructure. We already have our own CRM, AI agents, automated calculations, document processing, notifications, SLA management, and partial automation of internal workflows.

AI is not an experiment or an "extra feature" for us — it is a real operational layer of the business that already impacts request processing speed, communication quality, internal coordination, and team efficiency.

We are now at the point where we need to move beyond writing individual prompts and instead systematically build, stabilize, and scale the entire AI logic of the company.

## Why We're Hiring for This Role

We already have several agents in production, CRM integrations, structured outputs, routing logic, and automation across client-facing and internal workflows.

The challenge is that the system has grown complex: much of it relies on manual oversight, ad hoc agreements, and lives in one person's head. Any change to one agent can affect an adjacent process, and some of the logic is not yet sufficiently formalized, documented, or testable.

We need someone who will take ownership of the AI function: understand the current system, bring order to what exists, improve stability, eliminate gray areas, and help scale everything further — without chaos.

## What We Already Have

- Custom-built CRM
- Client-facing AI agents
- Scenarios for qualification, cost estimation, FAQ, summaries, checklists, and SLA
- Structured JSON responses and data-passing logic between agents and the system
- Automation of some internal and client-facing processes
- AI logic embedded in real business tasks, not in a test environment
- Active exploration of OpenClaw as a potential new branch of AI infrastructure development

For us, AI is part of the overall chain:
**User → Agent → Data collection & processing → CRM / internal logic → Team → Next step**

That's why it's not just about the quality of model responses — it's about the reliability of the entire system: no data loss, no broken logic, and agents that behave predictably.

## What You'll Be Doing

### Initially

- Study the current agents, their roles, and real-world usage scenarios
- Analyze existing prompts, logic, and weak points
- Identify duplication, chaos, instability, and hidden dependencies
- Structure the architecture of the AI logic layer
- Eliminate gray areas and unpredictable behavior
- Bring order to prompt versions, scenarios, and change tracking
- Build baseline test cases for key scenarios

### Ongoing

- Develop and refine agents for new business requirements
- Improve accuracy, stability, and predictability of responses
- Reduce hallucinations, unwanted improvisation, and failures at process handoffs
- Maintain JSON logic and data flow into the CRM
- Investigate real errors and edge cases reported by the team
- Collect user feedback and turn it into systematic improvements
- Contribute to new AI initiatives, including OpenClaw
- Gradually help transform the AI layer from a collection of solutions into a clear, maintainable system

### Dedicated Focus Area

- Build a testing ground for agents where conversations can be simulated and bulk-tested before deploying changes to production
- Create an environment for verifying agent behavior across standard, complex, and edge-case scenarios
- Develop automated prompt improvement pipelines based on accumulated data
- Identify contradictions, overlaps, and recurring patterns in the behavior of clients, managers, and agents
- Work with large volumes of already-collected conversations and scenarios to improve system behavior
- Participate in building HITL (human-in-the-loop) mechanics for agent fine-tuning and adjustment
- Move the AI layer from being manually configurable to gradually self-improving — under human oversight and with clear rules for making changes

## Who We're Looking For

We don't need someone who simply "writes prompts." We need someone who can build AI logic as a system and understand how it operates within a business.

### Required

- Experience building system prompts for real business use cases
- Understanding of how LLMs behave in practice: limitations, hallucinations, instruction-following, context management, few-shot techniques, etc.
- Experience with structured outputs, JSON, and scenarios where model responses feed into downstream systems
- Systems thinking and the ability to see the whole process, not just a single chat response
- Independence and willingness to navigate an already-running, live, and imperfect system
- Ability to distinguish whether a problem lies in the prompt, the data, the process, or the architecture itself

### Nice to Have

- Basic Python skills
- Understanding of APIs and integration logic
- Experience with Claude / OpenAI / other LLM APIs
- Experience with evals, AI testing, regression checks
- Experience with knowledge bases / RAG
- Experience in AI architecture, automation, or CRM workflows
- Experience building test environments, conversation simulators, or agent playgrounds
- Experience working with large conversation datasets, behavioral analysis, and HITL approaches
- Interest in OpenClaw or experience with similar approaches to AI context and memory management

### What Matters in a Person

- Quickly grasps context
- Takes ownership
- Thinks in systems, not fragments
- Can independently identify problems and propose solutions
- Understands how AI impacts processes downstream
- Can work not just with prompts, but with data, patterns, and overall system behavior quality
- Brings order and makes systems more reliable — not just "smarter"

## What We Offer

- Fully remote work
- Flexible engagement format
- Direct access to decision-makers
- A live system where much has already been built and is ready to be strengthened
- Significant influence over the development of the entire AI layer
- Opportunity to grow alongside the system and contribute to new initiatives
- Compensation discussed individually

We’re looking for a senior-level engineer to help design challenging, real-world coding tasks for evaluating AI coding agents.

This is not typical feature development. You’ll be creating engineering scenarios that require debugging, multi-step reasoning, and practical problem-solving inside Docker-based environments.

What you’ll do:
Design realistic engineering tasks (debugging, APIs, DB, systems)
Write clear instructions and expected solutions
Create tests/verifiers to validate completion
Ensure tasks require multi-step reasoning (not trivial problems)

Requirements:
Strong experience as a backend or full-stack engineer
Solid understanding of Docker and local dev environments
Experience with APIs, databases, and debugging complex systems
Ability to think like both an engineer and evaluator

Nice to have:
Experience with AI coding tools (ChatGPT, Copilot, etc.)
Experience writing coding challenges or technical assessments

Engagement:
Task-based work (2–5 hours per task)
Ongoing collaboration if quality is good

Please share relevant projects that showcase your ability to design non-trivial engineering tasks, work with Docker-based environments, and handle multi-step debugging or system-level challenges.
Looking for someone sharp, practical, and detail-oriented who understands real-world engineering problems.

JOB DESCRIPTION:

We are building a real-time financial analysis tool that processes live futures market data and delivers AI-powered analysis using a proprietary methodology. The system ingests structured market data, detects specific patterns, and outputs actionable analysis with suggested levels — all in real-time on a web interface. I need a senior Python/ML engineer to build V1 from architecture through deployment.

THE PROJECT
This tool combines three core systems: (1) a real-time data pipeline streaming 5-minute OHLCV bars from futures markets, (2) a RAG-based AI engine trained on a curated knowledge base of domain-specific content (~200–300 pages for V1), and (3) a React frontend displaying live charts with AI-generated analysis overlays. I provide all domain content and quality review. You build the engineering.

WHAT YOU WILL BUILD
• FastAPI backend with WebSocket support for real-time data streaming
• Integration with a financial data API (Databento) for live OHLCV futures data
• Custom pattern detection algorithms in Python (swing point identification, trendline detection, specific setup classification — detailed spec provided after screening)
• RAG pipeline: chunk, embed, and index ~200–300 pages of curated domain content into a vector database with metadata tagging
• LLM integration: assemble structured context, send to API (GPT-4o / Claude / DeepSeek), parse structured output
• React + TypeScript frontend with TradingView Lightweight Charts, AI overlay annotations, and analysis panel
• PostgreSQL + Redis for storage and real-time messaging
• Docker containerization and VPS deployment

MUST-HAVE SKILLS
• Python (advanced) — ~80% of the codebase
• FastAPI — REST + WebSocket endpoints
• Production RAG experience — at least one deployed RAG system (document chunking, embeddings, vector DB, prompt engineering, structured output parsing)
• PostgreSQL and Redis
• React + TypeScript

STRONG PLUS
• Experience with financial/market data APIs or real-time streaming data
• Technical analysis libraries (TA-Lib) or custom algorithm development
• Docker + deployment to VPS or cloud

TO APPLY — PLEASE INCLUDE:
• Describe one RAG system you built in production. What was the use case, which vector DB did you use, and how did you structure the data pipeline?

• Link to a relevant project or GitHub repo showing your Python/ML backend work.
• Your approach to the first 2 weeks of this project.
• Your fixed-price proposal for the V1 scope above.
Generic proposals without answering the above will not be reviewed.

• Describe a RAG system you built in production. What problem did it solve, what vector database did you use, and what was your chunking strategy?
• Have you worked with real-time data streaming (WebSockets, financial data feeds, or similar)? If yes, describe briefly.
• Link to a GitHub repo or portfolio project that demonstrates your Python backend + ML/AI integration skills.
• What would be your approach for the first 2 weeks of this project?
• What is your availability over the next 10 weeks, and are you currently committed to other projects?

We are seeking an experienced Pine Script developer to create a custom alert code for TradingView. The ideal candidate will have a strong understanding of trading strategies and the ability to write efficient, effective scripts that meet specific alert criteria. You will work closely with us to understand the setup and ensure that the alerts function correctly. If you have a proven track record of working with TradingView and can deliver quality scripts, we would love to hear from you. Please provide examples of previous work with Pine Script.

• Describe your recent experience with similar projects
• Please list any certifications related to this project
• Describe your approach to testing and improving QA

We are looking for an experienced Python backend developer to help extend and maintain a data processing system.

The backend is built using Python and an asynchronous web framework. The work involves implementing new data integrations, building API endpoints, and contributing to data pipeline functionality.

**Responsibilities:**

* Develop and extend backend services using Python
* Implement new data source and destination integrations
* Build and maintain REST API endpoints
* Work on data processing logic (batch and/or streaming)
* Write clean, maintainable, and well-structured code
* Collaborate on system design and integration strategies

**Requirements:**

* Strong experience with Python backend development
* Experience with async frameworks (e.g., Tornado, FastAPI, asyncio)
* Experience building or working with data pipelines / ETL systems
* Solid understanding of REST APIs and backend architecture
* Experience integrating third-party APIs or external data sources
* Ability to work independently and communicate clearly

**Nice to Have:**

* Experience with streaming systems
* Familiarity with databases (PostgreSQL, BigQuery, Snowflake, etc.)
* Experience with scalable backend systems

**Project Scope:**

* Initial task: Add a new data integration (source or destination)
* Potential for ongoing work depending on performance

**To Apply:**
Please include:

1. Relevant Python/backend projects
2. Experience with data pipelines or ETL systems
3. Experience with async frameworks
4. Brief explanation of how you would design and implement a new data integration

We are looking for someone practical and hands-on, with real project experience.

📝 JOB POST

i’m looking for a high-level match analyst to break down a singles pickleball match using pure analytics and data-driven insights.

this is not a general review — i want detailed tracking and pattern identification.

⸻

📊 match context
• i was up 11-1 in the first set
• ended up losing the set and match
• i want to identify exactly where and why the momentum shifted

⸻

🎯 scope of work

1. return analysis (primary focus)

track every return:
• placement (crosscourt, middle, down the line)
• depth (deep, mid, short)
• point outcome (won / lost / neutral)

👉 goal:
identify which return patterns led to winning vs losing points

⸻

2. shot selection breakdown

analyze key rallies:
• what shot was chosen
• whether it was high % or low %
• offensive vs defensive decisions

👉 goal:
identify where points were won vs given away

⸻

3. momentum shift identification
• pinpoint the exact score range where the match flipped
• identify what changed:
• return quality
• positioning
• decision-making
• shot tolerance

⸻

4. point construction patterns
• analyze first 3 shots and rally structure
• identify:
• winning patterns to repeat
• losing patterns to eliminate

⸻

📈 deliverables
• structured breakdown of:
• return outcomes
• shot selection patterns
• summary of:
• winning vs losing tendencies
• clear, actionable adjustments

bonus:
• screen-recorded walkthrough of key points

⸻

🧠 objective

understand:
• exactly why i lost after leading 11-1
• how to close matches consistently at a high level

⸻

⚡ ideal candidate
• strong understanding of high-level singles pickleball
• comfortable with data tracking + analytics
• able to provide clear, honest, actionable feedback

⸻

🔒 important

no fluff.
no generic advice.

i’m looking for pure analytics + real insights that translate to winning

TL;DR

I have a large dataset of U.S. commercial property addresses. I need a Python tool that enriches each address with tenant information, classifies the building by type (medical office, rehab facility, behavioral health, senior housing, hospital, etc.), and outputs a fully classified XLSX spreadsheet. An AI/LLM step handles ambiguous tenant names. I supply all necessary API keys and cover all usage costs. Milestone 1 is a working demo on a sample batch; Milestone 2 is the full dataset.

——————————————————

PROJECT OVERVIEW

I'm a healthcare commercial real estate investor. I have a large spreadsheet of U.S. property addresses, and I need each one researched and classified into one of the following categories:

• TARGET (Include): Medical Office Building (MOB), Inpatient Rehab Facility (IRF), Behavioral Health Facility
• CLASSIFY BUT EXCLUDE: Senior Housing / Assisted Living, Hospital Campus (unless freestanding outpatient building), Professional Office (under 50% healthcare tenancy), Retail (under 50% healthcare tenancy)

Important: Every address must remain in the final spreadsheet regardless of classification. "EXCLUDE" means the property is labeled as excluded from my target list — it does NOT mean the row is removed. I need to see every property and its classification.

Accuracy is critical. I need to know exactly what type of building each address is, how many tenants are in it, and what percentage are healthcare-related.

——————————————————

WHAT THE TOOL MUST DO (PER ADDRESS)

For each address in the dataset, the tool performs these enrichment steps in sequence:

Step 1: Location and Business Lookup
Identify the business name(s), property type, and any other businesses at or near the location. The goal is a baseline understanding of what exists at the address. You may propose the best data sources and approach for this step — I will provide any necessary API keys.

Step 2: Building Directory / Tenant List Research
Research each address to find a building directory or tenant list. The goal is to identify every tenant or suite in the building. Suggested search approaches:
• "[full address] building directory"
• "[full address] tenant list"
• "[full address] suite directory"
• "[building name] tenants"
Check common directory sources: LoopNet, CommercialCafe, officespace.com, building-specific websites.

Step 3: Tenant Identification and Classification
Extract every unique business/tenant name found at the address. For each tenant, classify as healthcare or non-healthcare.

Healthcare examples: physician practices (any specialty), surgical centers, imaging/radiology, PT/OT/speech therapy, dental, optometry, urgent care, behavioral health/psychiatry/counseling, pharmacy, lab, dialysis, OB/GYN, pediatrics, pain management, podiatry, chiropractic, allergy/immunology, home health agency offices, etc. These are common examples, not an exhaustive list — use judgment for tenant types not explicitly listed.

Non-healthcare examples: restaurants, cafes, gyms, law offices, CPA firms, insurance agencies, real estate offices, salons/spas, schools, retail, banks, tech companies, marketing agencies, etc.

Step 4: Non-Clinical Business Detection
For every address, run targeted research specifically looking for non-clinical businesses that might not appear in a building directory. Search for terms like: cafe, coffee, gym, fitness, restaurant, yoga, academy, school, salon, law, CPA, insurance, retail. Log every non-clinical hit — these affect the healthcare tenancy percentage calculation.

Step 5: Senior Housing Detection
This is the most important classification check. Many properties that look like healthcare facilities are actually senior housing. The tool must specifically look for these terms associated with each address:
• "memory care," "assisted living," "senior living," "independent living"
• "55+ community," "continuing care retirement," "skilled nursing facility"
If ANY of these terms are found associated with the address, classify as SENIOR HOUSING and label as EXCLUDE.

Step 6: Satellite / Street View Image (Optional but Preferred)
If possible, retrieve a satellite image and/or street-level image for each address and save the image URL for manual review of ambiguous properties.

——————————————————

CLASSIFICATION DECISION LOGIC

The tool must classify each property using this exact priority order. Check from top to bottom and stop at the first match:

1. Senior Housing keywords found? → SENIOR HOUSING (EXCLUDE)
2. Hospital campus? (ER, ICU, inpatient beds, trauma center) → HOSPITAL CAMPUS (EXCLUDE, unless freestanding outpatient building on campus → flag for manual review)
3. Behavioral health focus? (psychiatric, substance abuse, mental health residential) → BEHAVIORAL HEALTH (INCLUDE)
4. Inpatient rehab? (Encompass Health, Kindred, Select Medical, PAM Health, etc.) → IRF (INCLUDE)
5. Healthcare tenancy 50% or above? → MEDICAL OFFICE BUILDING (INCLUDE) — this applies even if the building appears to be a retail property
6. Healthcare tenancy under 50%? → PROFESSIONAL OFFICE (EXCLUDE)
7. Primarily retail with under 50% healthcare tenancy? → RETAIL (EXCLUDE)
8. Insufficient data? → UNKNOWN (flag for manual review)

——————————————————

AI/LLM CLASSIFICATION STEP

Some tenant names are ambiguous (e.g., "Wellbridge" could be a rehab facility or a fitness club; "Compass" could be a behavioral health provider or a real estate brokerage). For these cases, the tool should use an AI/LLM API to classify the tenant as healthcare or non-healthcare, sending the tenant name along with any available context (address, co-tenants, research snippets) and returning a confidence score.

You may propose the LLM that works best for this type of classification task. I will provide API keys. If you have a strong opinion on which model performs best for this kind of work, I'd like to hear it in your proposal.

——————————————————

OUTPUT REQUIREMENTS

The tool must produce a single XLSX spreadsheet with the following columns:

• Address — Original input address
• Building Name — Name of the building or primary business
• Classification — MOB / IRF / Behavioral Health / Senior Housing / Hospital Campus / Professional Office / Retail / Mixed Use / Unknown
• Include or Exclude — INCLUDE (MOB, IRF, Behavioral) or EXCLUDE (all others) — this is a label only; all rows remain in the spreadsheet
• Total Tenants Found — Number of distinct tenants/businesses identified
• Healthcare Tenant Count — Number classified as healthcare
• Non-Healthcare Tenant Count — Number classified as non-healthcare
• Healthcare Tenancy % — Healthcare tenants / total tenants
• Healthcare Tenant List — Comma-separated healthcare tenant names
• Non-Healthcare Tenant List — Comma-separated non-healthcare tenant names
• Senior Housing Flags — "None" or list of senior housing keywords found
• Non-Clinical Keyword Hits — "None" or list of non-clinical keywords matched
• Confidence Score — High / Medium / Low based on data availability
• Map Link — Direct link to view the address on a map
• Notes / Flags — Any context flagged for manual review

——————————————————

TECHNICAL REQUIREMENTS

• Python 3.10+
• You propose the best data sources and APIs for location lookup, tenant research, and enrichment — I will provide all necessary API keys
• AI/LLM API for ambiguous tenant classification (I will provide the key for whichever model you recommend)
• Rate limiting: respect all API quotas with appropriate delays
• Checkpoint/resume: if the tool stops partway through the dataset, it must be able to resume where it left off without reprocessing earlier addresses
• Input: CSV with one column of addresses
• Output: XLSX with all columns above
• Logging: detailed log file showing what was researched and found per address
• Error handling: if an API call fails for one address, log it and move to the next (no crashing the full run)

Note: I will cover all API usage costs. Your fixed price should reflect only your development time.

——————————————————

DELIVERABLES

• Complete Python tool with clear documentation and comments
• requirements.txt listing all dependencies
• README with setup and run instructions
• Sample output: process 100 test addresses I provide and show me the results before the full run
• Final processed XLSX for all addresses in the dataset
• Log file from the full run

——————————————————

MILESTONE STRUCTURE

Milestone 1 (20% of budget): Working tool + sample output on 100 test addresses. I review accuracy before approving. If accuracy on the sample falls below 90%, we work together to tune the logic before proceeding to the full run.

Milestone 2 (80% of budget): Full run on the complete dataset, delivered as a complete XLSX meeting all output requirements.

——————————————————

API KEYS AND DATA SOURCES

I will provide all necessary API keys after hiring. You do not need to create your own accounts. I am open to your recommendation on which data sources and APIs will produce the most accurate results for this type of property research. I already have accounts set up with major cloud and AI platforms.

——————————————————

TO APPLY

In your proposal, please include:
• A brief description of your experience with location data, property research, or similar data enrichment projects
• Your recommended technical approach (which APIs/data sources you would use and why)
• An example of a similar data processing project you've completed
• Your estimated timeline to complete both milestones
• Your fixed price for the project
• Any questions you have about the requirements (I value candidates who ask thoughtful questions)

I'm looking for someone who will be thorough and accurate. This could definitely lead to ongoing work as well! Quality matters more than speed. Thank you!

We’re looking for a full-stack developer to resolve a small but critical bug affecting our web application.
This is a targeted fix, not a full project. The issue is isolated — we need someone who can step in, identify the problem quickly, and implement a clean solution without overengineering.
Scope of Work:
Debug and identify the root cause
Fix a minor frontend or backend issue
Ensure everything runs smoothly after the fix
Perform a quick validation test
Requirements:
Solid full-stack experience (JavaScript, APIs, or backend logic)
Strong debugging mindset
Ability to deliver fast and clean

I have a large dataset (approx. 2GB) in Parquet format that needs specific transformations and schema validation. Since the data is too large for standard Pandas processing, I need a PySpark script that I can run on my local cluster to prepare the data for our analytics team.

Core Tasks
Schema Enforcement: Load raw Parquet/CSV files and enforce a strict schema (handling nulls and data type casting).
Data Transformation: * De-duplicate records based on a composite key.
Perform complex aggregations (e.g., daily rolling averages).
Filter out "junk" data based on specific business logic.
Validation Report: Generate a summary of "bad records" that failed the validation checks.
Output: Save the final cleaned dataset back to Parquet with optimal partitioning.

What I Need
I need a Python + Playwright script that runs automatically every night at 8:00 PM. It reads 200 leads from a Google Sheet, navigates through two authenticated insurance websites, collects specific data from each lead, and writes results back to the Google Sheet — all without any human involvement.

• Have you automated Salesforce Lightning or a similar enterprise CRM before? Briefly describe the project.
• Confirm you have used Playwright in Python. Describe one specific challenge you solved — for example handling dynamic page loads or authenticated sessions.
• Give me your day-by-day timeline for this project.

We are looking for a Python developer to build a script using Plyvel to read and parse LevelDB data from Chromium-based browser storage. The role involves extracting, decoding, and structuring key-value data (including binary formats) into usable outputs like JSON or CSV.

Candidates should have strong Python skills, experience with LevelDB, and familiarity with browser storage structures or low-level data parsing. Deliverables include a working script and clear documentation.

• Describe your recent experience with similar projects
• Please list any certifications related to this project
• Include a link to your GitHub profile and/or website
• What frameworks have you worked with?
• Describe your approach to testing and improving QA

Overview
We are looking for a skilled and confident full stack developer who combines strong technical ability with excellent communication skills.

This role goes beyond development it requires someone who can actively participate in discussions, explain solutions clearly, and represent work professionally in client-facing situations.

Key Responsibilities
Develop and maintain web applications using React and Node.js or Python
Participate in client calls and contribute to technical conversations
Clearly communicate ideas, updates, and solutions
Collaborate effectively within a remote and fast-paced environment
Required Qualifications
Proven experience with React and either Node.js or Python
Strong spoken English with clear and confident communication
Comfortable engaging in video calls and client-facing interactions
Ability to explain technical concepts in a simple and structured way
Reliable, responsive, and available for scheduled meetings
Preferred Background
Experience working directly with clients or stakeholders
Freelance or consulting experience
Background in startups or fast moving teams
Application Requirements
To be considered, please include:

A brief introduction about your experience
Your core tech stack and years of experience
A short video (2–5 minutes) introducing yourself and walking through a project you’ve worked on
Work Approach
We value individuals who demonstrate:

Clear and effective communication
Reliability and accountability
Professionalism in all interactions
A proactive and solution-oriented mindset
Opportunity
This role has the potential to grow into a long-term collaboration for the right candidate.

Description:

We’re looking for an experienced Electrical Engineer to help transform an existing hardware system into a connected, intelligent device with advanced software and communication capabilities.

This role is ideal for someone who can bridge hardware, firmware, and connected software systems—owning the development of a smart, WiFi-enabled platform from device to cloud.

Key Responsibilities:

Modify and integrate electronics within an existing hardware system
Develop embedded systems with:
WiFi connectivity (secure, reliable communication)
Over-the-air (OTA) updates
Real-time monitoring and control
Enable remote device management and automation
Design or support backend systems for:
User profiles and customization
Data handling and device communication
Open API access for future integrations
Help define system architecture across hardware, firmware, and software
Assist in identifying and onboarding additional specialists (e.g., mobile, backend) if/when needed

Requirements:

Strong background in electrical engineering and embedded systems
Experience with microcontrollers (ESP32, STM32, or similar)
Proven experience building IoT or connected devices end-to-end
Familiarity with firmware, OTA updates, and communication protocols (MQTT, HTTP, etc.)
Experience contributing to or designing backend systems/APIs
Ability to work with and enhance existing hardware systems

Nice to Have:

Experience with cloud platforms (AWS, Firebase, etc.)
Experience with mobile apps or user-facing interfaces
Background in smart devices, consumer electronics, or health tech

Project Scope:
This is an opportunity to take a physical product and evolve it into a fully connected system. We’re looking for someone who can lead initial development and help scale the technical approach as the project grows.

To Apply:
Please include examples of relevant projects, especially those involving connected devices, firmware, and system architecture.

Need Help Building the AI Trading Bot connected to Interactive Brokers.
I will be writing out the trading strategy and using Claude AI to help write the code. This will be around public stocks and indices.

Looking to work with up and coming freelancers that have the full skill set and looking to build professional resume and profile.

I need the help building the framework around connecting the code to the platform, paper trading testing, and live launch.

I am not the best at the specific coding and setting up the connections so looking for someone to walk me through it so it is built correctly and I can follow and make adjustments.

Would appreciate experience using Interactive Brokers API.

We’re building an AI booking and payment agent that runs inside WhatsApp.

We need a contractor to implement a Gemini multimodal normalization layer inside our Python AWS Lambda backend.

The system takes WhatsApp inputs (text, images, voice notes, documents, links) and converts them into structured JSON outputs for downstream processing.

Work includes:

integrating the Gemini API
implementing multimodal processing pipelines
handling speech-to-text and image-to-text inputs
integrating the solution into our existing Lambda Python architecture


Required skills:

Python
AWS Lambda
Gemini API / LLM integrations
Multimodal processing
Speech-to-text / image-to-text pipelines
API integration

We are building a private enterprise AI platform that allows organizations (banks, legal firms, and government agencies) to deploy secure internal AI assistants trained on their proprietary documents.
The platform will run inside the customer’s cloud environment (primarily Microsoft Azure) and provide capabilities such as:
• Internal policy and governance assistants
• Legal knowledge assistants
• Financial research assistants
• Document intelligence and retrieval
• Secure corporate AI knowledge bases
This is not a simple chatbot project.
It is a production-grade enterprise AI system built using RAG architecture, vector search, and secure cloud deployment.
We are looking for experienced engineers to help build the core backend infrastructure.
________________________________________
Roles We Are Hiring
We are open to hiring:
• Individual specialists
• A small team
• Or a full-stack AI engineer who can cover multiple areas
The project requires expertise in:
• Backend engineering
• AI / LLM integration
• Cloud infrastructure
• Data ingestion pipelines
________________________________________
Core Technology Stack
The expected stack includes:
Layer Technology
Backend Python (FastAPI preferred)
LLM orchestration LiteLLM / LangChain
Vector database Qdrant
Storage Azure Blob Storage
Cache Redis
Compute Azure Container Apps / Docker
Ingestion Python pipelines
Security Azure AD authentication
Monitoring Prometheus / Grafana
Experience with Azure OpenAI or AWS Bedrock is a strong plus.


The system will power AI assistants such as:
• Bank Policy GPT (internal policy assistant)
• Legal Research GPT
• Regulatory Compliance Assistant
• Corporate Knowledge Assistant
The platform must support multiple knowledge libraries.
Example:
• Legal library
• Governance library
• Policy library
• Research library

Please include:
1. Relevant AI or RAG projects
2. Cloud deployment experience
3. Preferred stack
4. Estimated timeline
5. GitHub or portfolio

• How do you train proprietary Data Stacks to work effectively with external models for inference
• How do you ensure embeddings from RAG are highly performant, especially in enterprise environments
• How would you build your model to learn and generate high-quality MS Office documents eg PPT, XLS. DOC, PDF, , proposals, pitch decks etc
• Describe your recent experience with similar projects

What You’ll Be Doing
Working on web applications using React, Node.js, and/or Python
Joining client calls and contributing to technical discussions
Explaining ideas, progress, and solutions clearly
Collaborating in a remote, fast-paced environment
What We’re Looking For
Solid experience with React and either Node.js or Python
Strong spoken English and clear communication skills
Comfortable on video calls and client-facing situations
Able to think on the spot and explain technical concepts simply
Reliable and available for meetings when needed

**Project Title:** AI Agent Developer — Franchise Consulting Business (Airia Platform)

**Overview**
We are a franchise consulting firm (team of 2–5) looking to hire an experienced AI agent developer to build AI-powered automation agents on the Airia platform. Our work is client-facing and involves confidential information, so data governance and security experience is essential.

The following is an overview of our initial scope, but we hope to continue to build more agents within Airia moving forward.

**Agents to Build**

1. Client Onboarding Document Generator
Automates the creation of personalised onboarding document packs for new clients. Pulls client details from our CRM and populates templates including welcome emails, project planners, and potentially scheduled meetings (if possible).

2. Research & Franchise Intelligence Agent
Given a franchise brand or sector, this agent researches and compiles market data, competitor landscape, franchise fees, and key risks into a structured client-ready brief.

3. Brand Template Customisation Agent
Customises our master document templates for each new franchise brand we work with — including logo swapping, find-and-replace of brand names throughout documents, and updating dates. Must work across Word (.docx).

**Technical Requirements**
- Hands-on experience building agents on the Airia platform (required)
- Experience with document manipulation libraries (e.g. python-docx, PyMuPDF or similar)
- Familiarity with Google Drive, Outlook, Office365 or CRM integrations via MCP or API
- Strong understanding of data privacy, confidentiality controls, and secure handling of sensitive client information
- Experience working with professional services or consulting businesses preferred

**How to Apply**
Please include the following in your application:
1. Confirmation of experience with the Airia platform
2. Examples of AI agents or automation workflows you have built (especially document-related)
3. Brief explanation of how you would approach data confidentiality for a client-facing consulting business
4. Your proposed approach — hourly or fixed price

We will start with a small paid test task before committing to the full project.

am seeking an expert Automation Engineer to build a sophisticated n8n-based AI Agent that performs comprehensive property and owner underwriting. This agent will ingest leads from my website and produce a "Ready-to-Sell" investment memo that includes deep-dive OSINT (Open Source Intelligence) and market velocity data.

The "Deep-Dive" Analysis Requirements:
The agent must use Gemini 1.5 Flash (Free Tier) and headless browsing to scrape/analyze:

A. The "Velocity" Layer: Pull 6-month Days on Market (DOM) trends and Absorption Rates from Redfin/Zillow public trackers to prove neighborhood liquidity.

B. The "Risk & Permit" Layer: Navigate the Maricopa/Pinal County Accela portals to identify open permits, unpermitted additions, and the age of major systems (HVAC/Roof).

C. The "Owner Psychology" Layer: Perform automated Google Dorking to find the owner's LinkedIn/Facebook. Extract age, marital status, and career history to generate high-conversion Talking Points and rapport-building strategies.

D. The "Mortgage Equity" Layer: Scrape the County Recorder for the original Deed of Trust. Use a custom JS node to calculate the Estimated Mortgage Balance via amortization.

E. Triple-Exit Matrix: Comparative math for Wholesale, Fix & Flip, and Subject-To, including a "Hidden Exit" analysis (lot split potential based on local zoning).

Technical Stack:

n8n: Advanced workflows with sub-flow logic.

LLM: Gemini 1.5 Flash (via Google AI Studio).

Automation: Playwright/Puppeteer or Skyvern for government portal navigation.

Data: Integration with U.S. Census API for neighborhood demographics/crime.

The Deliverable:
A structured Markdown "Deal Memo" (No emojis/AI-isms) delivered to Slack/CRM, including a 1-10 Confidence Score for the deal.

How to Apply:
Respond with the word "Architect." 1. How would you scrape the Pinal County Recorder site for a "Deed of Trust" without using a paid API?
2. What is your method for ensuring the LLM doesn't hallucinate the mortgage balance?