Available for talks & collaborations
Principal Software Engineer @ Autodesk
I build backend platforms, AI systems, and open-source tools — and occasionally talk about what I learn at conferences and meetups. Based in Bangalore.
Writing
Notes on AI systems, cloud architecture, and software engineering.
Most RAG tutorials stop at chunk-embed-retrieve. Here is what the production gap actually looks like and how to close it.
Three approaches to giving LLMs access to external capabilities. Here is when to use each one and why the choice matters more than most people think.
After integrating three different LLM providers across two products, I got tired of rewriting the same patterns. Here is what I built and what the experience taught me about production AI systems.
Open Source
Open-source tools for building production AI systems.
Modular TypeScript library for building production-ready AI chat applications. Provides a unified interface over multiple LLM providers with built-in streaming, history management, and tool calling.
Express-based server adapter for ai-chat-toolkit. Exposes a REST and SSE API so any frontend can connect to your AI chat backend without boilerplate.
Drop-in React chat widget for ai-chat-toolkit-server. Zero-config embed with customisable theme, streaming responses, and mobile-friendly layout.
Speaking
Conference talks and meetup sessions on AI, cloud, and engineering.
Backend engineers who understand cloud architecture ship better systems. This talk covers event-driven architecture, serverless patterns, da…
Most RAG tutorials stop at "chunk text, embed, retrieve". This talk goes further — covering chunking strategies, metadata filtering, hybrid …
The AI tooling landscape is noisy. This talk cuts through the hype by comparing three popular approaches to giving LLMs access to external c…