Full Stack Developer

We are launching a specialized, high-stakes project: a real-time, conversational AI Search Engine tailored entirely for the Healthcare and Medical niche. This is a long-term, high-commitment engagement (30+ hours/week) for a senior architect who can bridge complex medical data retrieval with elite AI engineering. ​We are not building a simple wrapper. This platform must dynamically parse medical queries, fetch real-time clinical and academic data from validated sources, run advanced medical RAG pipelines with strict semantic verification, apply smart reasoning to filter out noise, and stream answers with precise, ironclad inline citations. ​What You Will Do: ​Architect an ultra-fast, streaming Next.js interface capable of rendering complex medical data, tabular clinical trial results, and interactive, multi-turn follow-up reasoning paths. ​Design a high-concurrency Python backend to handle parallel medical search query routing, clinical database querying, and real-time medical text parsing. ​Implement production-grade Medical RAG pipelines featuring multi-stage retrieval, semantic reranking, and domain-specific medical chunking. ​Utilize Model Context Protocol (MCP) to securely and cleanly interface between LLMs, medical vector databases, and external clinical APIs. ​Configure Smart Reasoning Models to accurately categorize search intent (e.g., distinguishing between patient-friendly explanations vs. deep clinical research syntax). ​Integrate secure Stripe payment infrastructure customized for tiered enterprise or provider subscriptions and usage-based API/token limits. ​Requirements: ​10+ years of full-stack engineering experience with a portfolio showing live, production-grade applications. ​Direct Healthcare AI Experience: Proven experience building medical data pipelines, handling high-stakes contextual datasets, or developing clinical search/retrieval tools. ​Absolute Autonomy: Capable of engineering complex asynchronous data flows, managing strict latency limits, and implementing rigorous hallucination safeguards. ​Availability: 30+ hours/week with tight, reliable response times (0–4 hours) during US business hours. ​Required Tech Stack: ​Frontend & Streaming: Next.js (App Router), TypeScript, Tailwind CSS, Vercel AI SDK. ​Core Systems: Node.js, NestJS, Python (FastAPI / Django). ​AI Search Architecture: Advanced RAG, Vector Databases (Pinecone, Weaviate, or pgvector), Cohere Rerank, MCP tools, and advanced reasoning models (OpenAI o-series, Claude 3.5 Sonnet, Gemini 1.5 Pro). ​Infrastructure: PostgreSQL, Redis (highly optimized for search caching), MongoDB, and native Stripe billing.