Fix RAG service & Voice Interaction Issues in Python-Based Healthcare Application
Budget: $25 – $50 USD
I need an experienced AI/ML developer to diagnose and resolve specific technical issues in a Python-based healthcare application that uses Retrieval-Augmented Generation (RAG) for content delivery and Vapi.ai for AI voice agent integration.
Specific Problems to Solve:
1. RAG System Issues:
- Vector search returning irrelevant clinical content chunks
- Embedding mismatches causing incorrect template retrieval
- Context window optimization needed for accurate responses
2. Voice Interaction Bugs:
- Speech interruption detection not triggering correctly
- Silence timeout thresholds causing premature disconnections
- Response retry logic failing after user interruptions
3. Content Validation:
- Output filtering not enforcing template structure
- Responses deviating from approved clinical script formats
Required Skills:
- Strong Python debugging capabilities
- Hands-on experience with LangChain, LlamaIndex, or similar RAG frameworks
- Experience with STT/TTS features using Vapi.ai
- Prompt engineering and LLM output structuring
Preferred Experience:
- HIPAA-compliant application development
- Healthcare data systems or EHR integration
What You'll Deliver:
- Root cause analysis document for each bug
- Fixed Python code with inline comments explaining changes
- Test results proving issue resolution
- Setup/deployment notes for future maintenance
To Apply, Please Include:
1. Brief description of similar RAG or voice AI debugging work
2. Your systematic approach to troubleshooting LLM applications
3. Current availability (hours per week)
4. Your hourly rate or fixed-price proposal
Specific Problems to Solve:
1. RAG System Issues:
- Vector search returning irrelevant clinical content chunks
- Embedding mismatches causing incorrect template retrieval
- Context window optimization needed for accurate responses
2. Voice Interaction Bugs:
- Speech interruption detection not triggering correctly
- Silence timeout thresholds causing premature disconnections
- Response retry logic failing after user interruptions
3. Content Validation:
- Output filtering not enforcing template structure
- Responses deviating from approved clinical script formats
Required Skills:
- Strong Python debugging capabilities
- Hands-on experience with LangChain, LlamaIndex, or similar RAG frameworks
- Experience with STT/TTS features using Vapi.ai
- Prompt engineering and LLM output structuring
Preferred Experience:
- HIPAA-compliant application development
- Healthcare data systems or EHR integration
What You'll Deliver:
- Root cause analysis document for each bug
- Fixed Python code with inline comments explaining changes
- Test results proving issue resolution
- Setup/deployment notes for future maintenance
To Apply, Please Include:
1. Brief description of similar RAG or voice AI debugging work
2. Your systematic approach to troubleshooting LLM applications
3. Current availability (hours per week)
4. Your hourly rate or fixed-price proposal