Head of DevRel at TrueFoundry — I work at the intersection of AI infrastructure and developer experience. Mostly thinking about agents, reliability, and what happens when LLM systems meet production.
Preparing for a world with no UI.
AgentBreak — chaos testing for AI agents. Break them intentionally so they don't break when it matters.
eyeroll — AI eyes that roll through video footage. Watches screen recordings, Loom videos, and YouTube links to help coding agents fix bugs and build features.
VoiceAudition — the casting director for AI voice agents. 697+ voices across 9 TTS providers with semantic search and auditions.
tfy-gateway-skills & tfy-deploy-skills — agent skills for deploying and routing on TrueFoundry's platform.
Cognita — led and maintained TrueFoundry's open-source RAG framework.
Preclinical — open-source platform for adversarial safety testing of healthcare AI agents. Built around TriageBench, a benchmark that evaluates whether AI agents can correctly triage patients into emergency, clinician, or home care — and catches the dangerous errors where agents under-triage emergencies. Paper in collaboration with UCLA, coming soon.
- Voice agents that respond in under 500ms — latency is the product
- Agents that survive the real world, not just the demo
- The infrastructure layer that makes all of this actually work in production
- Clarity over anything else
- Speaking at events and conferences
- Judging and sponsoring AI hackathons
- Building a builder-first AI community — join the discord
- Building something together? → mnvsk97@gmail.com
- Collaborating with TrueFoundry? → sai@truefoundry.com
- In SF? Let's grab coffee ☕