Engineer at Nous Research, based in Sydney. Previously 10+ years at Atlassian.
I work on the messy middle of agentic engineering, making agent workflows inspectable, testable and dependable enough to trust.
- The Unbusy Co - Founder. We find the admin drag in growing teams and build workflows that run themselves.
- ShakedownKit - Ultralight backpacking gear management.
- ngxcalvin.com - Writing on AI, engineering, and adventure.
- Agent Swarm - Many agents on the problem, one answer back.
- Agent Profile Benchmark - Reproducible evals for AI agent profiles on real tasks.
- Reasoning Benchmark - Does the model understand the question, or just the words?
- Skill Suitcase - Portable agent skills across toolchains.
- Hermes Discord Clarify - Readable multiple-choice prompts for Hermes in Discord.
- Artshelf - Accountable retention for the files agents leave behind.
- Momentum - CLI for autonomous repo work with local verification.
Outside the terminal: mountaineering, hiking, surfing, strength training, and pack-weight spreadsheets.
