Full-stack developer building fast, well-typed tools. Lately, ways to make LLM applications spend fewer tokens.
I'm a full-stack developer with a soft spot for data structures, algorithms, and systems that don't waste resources whether that's a database query or a prompt sent to an LLM.
My current focus is developer tooling for AI applications: measuring, trimming, and controlling how many tokens a system actually needs to spend. Outside of that, I build the usual full-stack range: APIs, data-backed apps, and the occasional weekend script.
A production-grade Python framework for measuring and reducing LLM token usage, built with a zero-dependency core, an optional tiktoken backend for exact counting, and a fully typed, tested, CI-linted codebase (mypy --strict, ruff, pytest).
Marketing site and brand for an INR ↔ crypto exchange service built around a live rates dashboard (USDT, BTC, ETH, LTC), an animated stats strip, and a community reviews section for a 4,000+ member Discord-based exchange.
Always interested in hearing about new opportunities, collaborations, or just talking shop about LLM tooling.