Product platform
Frontend and backend foundations for tinnitus education, programs, and patient-facing guidance.
Case study
Tinni-Care: healthcare platform with RAG-based product assistant.
Tinni-Care is a healthcare platform co-built with an ENT physician. As CTO and co-founder, I lead architecture, product engineering, and AI integration across the platform.
Project context
Project context
The product needed to combine expert-led medical content, patient-facing product UX, scalable platform features, and AI-assisted guidance without treating the assistant as a generic chatbot.
Challenge
Challenge
The difficult part was not only building a web app. The challenge was creating a product architecture where expert content, patient-facing guidance, platform workflows, and AI retrieval could work together in a maintainable way. The assistant needed to be useful inside the product, grounded in relevant content, and presented with careful UX boundaries.
Approach
I shaped the system around four connected layers:
Content workflows
Structures for expert-led material and product content that could later support retrieval and assistant behavior.
Retrieval and AI API
RAG-oriented data flow with vector search and assistant responses connected to product context.
UX and safety constraints
Interface patterns, assistant boundaries, and product copy that help users understand what the assistant can and cannot do.
Architecture notes
- Next.js product frontend
- Node.js backend/API layer
- PostgreSQL data model
- pgvector for retrieval workflows
- LLM assistant integration
- Content and data-flow design
- Deployment and platform infrastructure
Visual proof
Assistant flow and architecture shape.
Video preview: the Tinni-Care website assistant answering a tinnitus question from expert content.
Expert content
Medical and expert tinnitus material prepared for product use.
Retrieval index
Structured content and vector retrieval workflow.
AI API
Assistant responses connected to platform context.
Product assistant UI
User-facing assistant experience with clear boundaries.
Outcome
Outcome
The result is a healthcare product platform where architecture, product engineering, expert content workflows, and AI assistant integration are connected instead of treated as separate tracks.
What this proves
Relevant for similar client work.
- A senior technical owner for an AI-enabled product.
- Full-stack product architecture and implementation.
- RAG or LLM features connected to real content.
- Product UX that makes AI features understandable and useful.
Next step
Send a concise project brief.
Share what you want to build or improve, where the project stands, and what kind of senior technical help you need.