Headline

Full Stack Engineer - Python | FastAPI | React | GenAI | RAG | LangChain | SQL | Docker | Cloud | Backend Developer | API Engineer

About

I build backend systems and production AI features that remove friction from clinical workflows — the kind of software that doctors and coordinators actually depend on, not demos. In 4 years at a healthcare product company, I've worked across the full stack: FastAPI backends, React TypeScript frontends, PostgreSQL query optimisation, and GCP Cloud Run deployments. But increasingly my focus has shifted to where engineering meets AI. Most recently I built a RAG-based clinical document platform that takes a 20+ minute manual document review down to under 30 seconds. The hard part wasn't picking LangChain — it was figuring out that generic token-based chunking broke citation accuracy for medical documents, and designing a section-aware chunking strategy around clinical headers (DIAGNOSIS, HISTORY, MEDICATIONS) that got precision from ~61% to ~89%. I also built an LLM-powered data validation engine where the model infers column types from unlabelled CSVs, detects medically impossible values, and dynamically generates Pydantic validation rules — cutting manual schema mapping entirely. On the infrastructure side, I've taken SQL stored procedures from 60s to 12s, improved API response times 30–40%, and led a platform migration from a layered monolith to domain-driven design. What I'm looking for next: a role where the scale is bigger, the engineering team is stronger, and the AI problems are harder. If you're building something at the intersection of reliable software engineering and production AI, I'd like to hear about it. 🛠 Stack: Python · FastAPI · LangChain · ChromaDB · React · TypeScript · SQL · GCP · Redis · Celery · Docker

Location

Pune/Pimpri-Chinchwad Area

AI stack

codex

cursor

opencode

Work experience

Product Engineer at IKS Health, May 2022 - Present · 4 yrs 3 mos, Navi Mumbai, Building backend systems and AI features for enterprise healthcare software used by clinical operations teams across India. • Reduced SQL stored procedure execution time from 60s → 10–12s by redesigning composite indexes and eliminating correlated subqueries; improved report generation speed by 40–70% across the platform • Improved API response times 30–40% by refactoring async pipelines, introducing Redis caching, and eliminating N+1 query patterns across critical clinical workflow endpoints • Built a RAG-based clinical document Q&A platform (MedDoc AI) using LangChain, ChromaDB, FastAPI, and GCP Cloud Run — reduced manual document review from 20+ min to under 30 sec; designed section-aware chunking strategy that improved citation precision from ~61% to ~89% • Built an LLM-powered data validation engine (SmartPipeline) using Gemini API and Pandas — eliminated manual schema mapping from unlabelled CSVs; enabled plain-English dataset queries via sandboxed NL-to-Pandas expression evaluation • Designed and implemented JWT + RBAC + refresh-token rotation authentication for multi-tenant healthcare applications with Active Directory integration • Built 100k+ row data ingestion pipeline using Pydantic v2 validation, chunked bulk inserts, and structured per-row error reporting — eliminated manual data correction cycles • Reduced production incident MTTR by 50% through structured JSON logging, correlation IDs, and GCP Cloud Logging observability improvements • Led platform architecture evolution from layered monolith to Domain-Driven Design — defined domain boundaries, API contracts, and frontend feature modules across the engineering team • Designed React + TypeScript dashboards with WebSocket real-time updates and advanced filtering, replacing manual Excel-driven workflows for clinical operations teams

Education

Postgraduate Diploma, Big Data Analytics, Centre for Development of Advanced Computing (C-DAC)

Bachelor's degree, Mechanical Engineering, Padmabhooshan Vasantdada Patil Institute of Technology, Bavdhan