Hey, I'm Tarun.
engineering systems, AI infrastructure, and developer tools.
I care about systems that stay correct under load: clear interfaces, intentional trade-offs, and failure modes you can reason about.
Most of my time goes into backend services, execution pipelines, and the glue that makes AI features reliable in production.
Projects
Articles & case studies
Why Context Windows Are The Wrong Abstraction
Article · 2026
My thoughts on long-term memory for AI systems, and why stuffing the prompt is not a strategy.
Read
Context Memory Service
Case Study · 2025
A retrieval layer that stores, ranks, and injects long-term context for multi-agent systems without blowing the prompt window.
Read
Building an execution engine that can fail safely
Article · 2025
Leases, idempotency keys, and why 'retry' is not a recovery strategy by itself.
Read