Himansh Raj
Applied AI Engineer
I build the layer underneath the models: model gateways, agent runtimes, and inference that runs on your own machine.
At Colate I built the multi-provider LLM gateway behind Cana: one shared request/event representation translating OpenAI, Anthropic and Gemini formats, with streaming, tool calls and multimodal inputs. Prompt caching and token accounting took it to 6.05B cached input tokens, 53% input-token reuse and 35% lower inference latency. I also built its multi-agent orchestration and persistent AWS/Kubernetes runtimes.
Alongside that I research agentic systems at the MISN Lab, IIT Delhi, and train small models from scratch to understand how they fail.
Maya OS
maya.himansh.in ↗A local-first AI operating system for developers. One command boots an OpenAI and Anthropic compatible gateway over cloud and local GGUF models, real-PTY agent teams, a message bus and a dashboard. Eight packages, MIT, shipped to npm.
npm i -g maya-os