Building trustworthy infrastructure for autonomous AI systems.
I design distributed systems for production today, and I research the infrastructure required for reliable agentic AI — memory, observability, governance, orchestration, and evaluation.
Works with your stack — and the rest of it too
About the researcher
- 01Networking
- 02DevOps
- 03Site reliability
- 04Security engineering
- 05Infrastructure architecture
- 06Agentic systems
I began my career in networking before moving through DevOps, site reliability, and security engineering into infrastructure architecture. Over the past decade I have designed and operated distributed systems for production environments, increasingly working at the boundary between infrastructure, security, and automation.
That progression shaped my interest in autonomous systems: if infrastructure can already be automated, what happens when the systems responsible for operating it can reason, adapt, and act?
My current work explores the infrastructure required to make those systems reliable — memory, observability, governance, orchestration, and evaluation. I spent a decade learning how complex production systems fail; the open question I am now investigating is what happens when those systems become autonomous.
Read the full research visionResearch systems: questions with running code.
Recent articles.
Notes from the workbench.
August 2026 — Building memory isn't enough
The harder problem is deciding what should be remembered, and what should be allowed to fade.
July 2026 — Uncertainty is information, not a bug
The most interesting failures are not when an agent is wrong, but when it is uncertain and acts anyway.
Open source, mapped to research areas.
Each repository implements one research area. Star and fork counts are pulled directly from the GitHub API.
A decade of production systems, as evidence for the research.
Distributed systems, platform engineering, and security work across startups and enterprises. The professional record is context for the research direction, not the product.
Hands-on help with agentic AI infrastructure.
Architecture review, security baselines, and implementation support for teams building autonomous systems.
Talk to an engineer, not a sales bot.
Tell us about your stack and what you're trying to automate. We'll reply within one business day with a concrete next step — pilot, architecture review, or a quick async answer.