Build Python orchestrator with Streamlit UI for parallel multi-LLM querying
Budget: $600 – $1,000 USD
We need a senior Python developer to build a custom internal research tool for our healthcare AI consultancy. The tool runs 75 standardized queries across six AI platforms in parallel, captures structured responses and citations, and writes results to an Excel template for analysis.
What you're building
A Python application with three components:
Six API adapters: OpenAI (ChatGPT), Anthropic (Claude), Perplexity, Google (Gemini), xAI (Grok), and SerpAPI (for Google AI Overviews). All hit the official APIs and return a normalized response object.
Async orchestrator that fires all 75 queries in parallel across the six platforms, handles retries and rate limits, tags client domain and competitor mentions in real time, and writes results to a provided Excel template.
Streamlit UI for click-to-run operation, with engagement metadata input, live progress monitoring, results preview, plus Google Drive upload and Slack notification integrations on completion.
Deliverables
Six platform adapter modules with normalized response interface
Async orchestrator with error handling and retry logic
Excel writer (we provide the template)
Streamlit dashboard
Google Drive and Slack integrations
README, setup guide, and operator documentation
Test suite covering 70%+ of code paths
All code in a GitHub private repo owned by us
What we provide at kickoff
Detailed architecture document and UI specification
AHS Excel scan template, production-ready
Full 75-query taxonomy
All six API keys, provisioned and shared via 1Password
Sample completed scans for reference
Direct Slack access to the project lead
Timeline and process
10 business days from kickoff to delivery
Fixed price, paid in three milestones (30% / 40% / 30%)
$500 bonus if delivered on or before business day 8
Daily commits to the repo required
Weekly Friday sync at 10am ET
Mutual NDA signed before access to query taxonomy
Required skills
5+ years Python with strong asyncio experience
Hands-on experience with at least 3 of: OpenAI API, Anthropic API, Perplexity Sonar API, Google Gemini API, xAI Grok API, SerpAPI
openpyxl for Excel manipulation
Streamlit for UI
Google Cloud OAuth for Drive integration
Slack API for notifications
Strong Git/GitHub workflow
Out of scope
Query design, AI scoring synthesis, deck generation, mobile, and multi-user authentication. This is one focused tool with a clear endpoint.
Please answer in your proposal
Walk us through a recent Python project that integrated 3+ third-party APIs. What broke and how did you fix it?
Which of the six APIs in scope have you used in production?
What's your approach to rate limits and transient failures across async API calls?
What scope questions do you have before bidding?
Not a fit if
You've never integrated more than two LLM APIs, you're uncomfortable with fixed-price contracts, or you can't commit to daily commits on a 10-business-day timeline.
What you're building
A Python application with three components:
Six API adapters: OpenAI (ChatGPT), Anthropic (Claude), Perplexity, Google (Gemini), xAI (Grok), and SerpAPI (for Google AI Overviews). All hit the official APIs and return a normalized response object.
Async orchestrator that fires all 75 queries in parallel across the six platforms, handles retries and rate limits, tags client domain and competitor mentions in real time, and writes results to a provided Excel template.
Streamlit UI for click-to-run operation, with engagement metadata input, live progress monitoring, results preview, plus Google Drive upload and Slack notification integrations on completion.
Deliverables
Six platform adapter modules with normalized response interface
Async orchestrator with error handling and retry logic
Excel writer (we provide the template)
Streamlit dashboard
Google Drive and Slack integrations
README, setup guide, and operator documentation
Test suite covering 70%+ of code paths
All code in a GitHub private repo owned by us
What we provide at kickoff
Detailed architecture document and UI specification
AHS Excel scan template, production-ready
Full 75-query taxonomy
All six API keys, provisioned and shared via 1Password
Sample completed scans for reference
Direct Slack access to the project lead
Timeline and process
10 business days from kickoff to delivery
Fixed price, paid in three milestones (30% / 40% / 30%)
$500 bonus if delivered on or before business day 8
Daily commits to the repo required
Weekly Friday sync at 10am ET
Mutual NDA signed before access to query taxonomy
Required skills
5+ years Python with strong asyncio experience
Hands-on experience with at least 3 of: OpenAI API, Anthropic API, Perplexity Sonar API, Google Gemini API, xAI Grok API, SerpAPI
openpyxl for Excel manipulation
Streamlit for UI
Google Cloud OAuth for Drive integration
Slack API for notifications
Strong Git/GitHub workflow
Out of scope
Query design, AI scoring synthesis, deck generation, mobile, and multi-user authentication. This is one focused tool with a clear endpoint.
Please answer in your proposal
Walk us through a recent Python project that integrated 3+ third-party APIs. What broke and how did you fix it?
Which of the six APIs in scope have you used in production?
What's your approach to rate limits and transient failures across async API calls?
What scope questions do you have before bidding?
Not a fit if
You've never integrated more than two LLM APIs, you're uncomfortable with fixed-price contracts, or you can't commit to daily commits on a 10-business-day timeline.