ML Engineer. Researcher.
I specialize in machine learning engineering and systems. Currently, I'm at Scale AI, building infrastructure for the agentic web. My open-source work focuses on the "plumbing" of this new ecosystem—creating protocols, auth frameworks, and tools that help agents communicate and operate reliably.
- Repowire (GitHub) – Mesh network enabling Claude Code and OpenCode sessions to communicate in real-time across repos.
- FastHarness (GitHub) – Decorator-based framework for exposing Claude SDK agents through Google's A2A protocol.
- A2A Registry (GitHub) – A community-driven directory of live, hosted AI agents using the A2A Protocol.
- Phlow (GitHub) – JWT-based authentication framework for secure agent-to-agent communication.
- CloudRegion (GitHub) – Intuitive cloud region mapping across AWS, Azure, and GCP.
- Modalkit (GitHub) – Production-ready framework for deploying ML models on Modal.
- Tracelet (GitHub) – Unified experiment tracking across MLflow, WandB, and ClearML with zero code changes.
- Torale (torale.ai) – AI-powered web monitoring platform tracking meaningful changes across the internet.
- Feed Forward Podcast (Listen) – Co-hosting discussions on the AI landscape.