Backends for AI agents
- Java
- Spring Boot
- MCP
- REST
Task, execution and planning APIs behind production AI agents, and a feedback loop that turns observed agent failures into new tools.
Read the case studyHello.
I’m Om
Ten years building secure, multi-tenant enterprise software, and now the backends behind AI agents. Off the clock I build my own apps, run my own servers and write about both.

Grew from engineer to lead across three product areas. Now leading the middleware and API layer of an enterprise risk platform and building backends for AI agents.
Led a module of an identity product, from requirements to release.
I like systems that are secure, fast and easy to operate. At work that means search, event pipelines and API platforms, and leading the engineers who build them. Lately it also means giving AI agents the tools they need to be useful.
Outside work I build the OneLifeStack apps, run a small homelab and a VPS, and share what I learn on my blog. I’m based in Pune, India.
Task, execution and planning APIs behind production AI agents, and a feedback loop that turns observed agent failures into new tools.
Read the case studyA search and aggregation layer plus Kafka-driven pipelines over large multi-tenant datasets, with strict per-customer scoping and a performance program.
Read the case studyThree Spring Boot codebases bootstrapped from nothing, two of them taken to production, with multi-tenant isolation, auth and Kubernetes delivery.
Read the case studyA family of native Android apps that cooperate on the device, plus the web, backend and self-hosted infrastructure around them.
Read the case studyWhat routing and tunnelling mean, why reverse tunnels let you host from behind a home router, and how hostname routing sends traffic to the
A short, practical intro to tmux: what it is, why it saves remote work, and the handful of keys and settings worth learning first.
Hiring, a collaboration, or a question about something I wrote? Send a note, or email me directly.