Harsh Prajapati

I'm a backend-first Python engineer with 3 years of experience shipping production systems with FastAPI, Django REST Framework, PostgreSQL, Docker and AWS EC2. For the past 2 years I've applied that backend foundation to GenAI: RAG pipelines, LLM-powered APIs and agentic workflows with LangChain and LangGraph. I'm based in India (IST, UTC+5:30), have worked remotely on contract for a Canadian company, and I'm comfortable owning work end to end in async, distributed teams.

Recent work: I independently designed and deployed a modular FastAPI platform on AWS EC2 that auto-mounts new tool backends at startup. It hosts 25+ lead-gen tools for ~15-20K users/day, with Redis rate limiting, usage-tracking middleware and a CI/CD template that cut new-tool setup time by 60%. I also built a live AI real estate platform end to end (OpenAI intent routing over live MLS listings), and contributed to a compliance-review AI platform with JWT/RBAC APIs, background jobs and an agentic audit workflow.

On the GenAI side, I built an async, idempotent RAG pipeline over SEC EDGAR filings: PostgreSQL as source of truth, Qdrant as a rebuildable vector index, and hybrid retrieval (dense + BM25 + Reciprocal Rank Fusion) that scores 0.85 nDCG and 0.96 MRR on a 25-query eval. Earlier, I built Elasticsearch and Pinecone search and OpenAI summarization pipelines, plus Flutter and React front ends when a product needed them.

Looking for remote Backend, GenAI/Applied AI or Python roles. Stack: Python, FastAPI, Django, PostgreSQL, MySQL, Redis, SQLAlchemy, Celery, Docker, GitHub Actions, OpenAI API, LangChain, LangGraph, RAG, Qdrant, Pinecone, Elasticsearch, React, Next.js.