AI Infrastructure / Backend Engineer — India · Remote
Sai Krishna Podduturi
I build the unglamorous systems that make AI actually usable in production — gateways, auth, metering, and the Kubernetes plumbing underneath.
I'm an engineer who lives at the seam between applications and infrastructure. Most recently I've been building self-hosted AI platforms — putting a clean, metered, multi-tenant API in front of open models so teams can ship without handing their data (or their budget) to someone else's cloud.
I care about the parts users never see: predictable latency, sane rate limits, auth that doesn't leak, cost you can actually attribute, and deploys that don't page you at 3 a.m. I run my own k3s cluster end to end — ingress, TLS, Postgres, CI → registry → rollout — so the things I build are things I also operate.
-
01
AI Gateway 2025—
An OpenAI-compatible proxy in front of self-hosted LLMs (Ollama). Per-key authentication, rate limiting, credit & usage metering, and an admin API — so a whole team can share local models with real guardrails instead of an unmetered free-for-all.
-
02
Aarogya kinveetech
Health-tech platform work under kinveetech — backend services and the infrastructure behind them.
-
03
Your next highlight
Add a third project here — ideally one with users, scale, or a number attached.
- Languages
- Go · Python · SQL · Bash /* TODO: adjust */
- Infrastructure
- Kubernetes (k3s) · Docker · cert-manager · NGINX Ingress · Cloudflare · GitHub Actions
- Data
- PostgreSQL · Redis /* TODO */
- AI / ML
- LLM serving · Ollama · OpenAI-compatible APIs · RAG · prompt & tool orchestration
- Practices
- API design · observability · CI/CD · infra-as-code · on-call ownership