// arpit.chauhan
Building observable backend systems, applied AI infrastructure, and fault-tolerant distributed services.
Independent engineering work spanning multi-agent pipelines, lease-based job runners, layout-aware retrieval, and edge auth — designed and operated end to end.
// core areas
- Backend Systems
- Distributed Systems
- Applied AI Infrastructure
- Observability
// primary technologies
Python · FastAPI · PostgreSQL · Redis
Docker · AWS · Prometheus · Grafana · Tempo
Google ADK · TypeScript
// selected work
Projects
- 01Blogify AIPython · FastAPI · Google ADK · Postgres
Multi-agent content pipeline with sandboxed secrets and cost telemetry.
5 agents - 02Distributed System BasicsPython · FastAPI · PostgreSQL · Redis
Lease-fenced async job runner with Postgres as truth, Redis as transport, and a reaper for stale work.
18.86s p95 terminal - 03Pinpoint PDFPython · FastAPI · Pinecone · PyMuPDF
Layout-aware RAG with multi-query, RRF, and a cross-encoder rerank stage.
3-stage retrieval - 04E-Library REST APITypeScript · Express · Postgres · Redis
RBAC-protected Express/TS API with Redis cache-aside and Prometheus SLIs.
< 50ms read p95 - 05Blogify EdgeTypeScript · Hono · Cloudflare Workers · Web Crypto
Hono on Workers with OAuth2 PKCE and Web Crypto-signed sessions.
~2–3× faster sign-in
// currently
Exploring systems programming in C — concurrency primitives, process isolation, and event-driven I/O. Continuing applied AI infrastructure work focused on observability, cost, and safe agent execution. Recognized in AI for Bharat top 30; previously collaborated on WalkSense.
// contact
Get in touch
Best for: backend systems, applied AI infrastructure, and platform engineering collaborations.