The Algorithm
Notes on building AI systems that actually work.
What I've learned building AI systems across healthcare, finance, ecommerce, and government — the architecture decisions, design tradeoffs, and organizational dynamics that determine whether an AI project succeeds or stalls.
Latest
I Want to See My Agents Work
If people remain responsible for work done by AI agents, they need to be able to see the work in progress and inspect the record it leaves behind.
Read
Guardrails Aren't Containment
OpenAI turned down its models' cyber refusals on purpose, to measure real capability. Then one of those models found an unknown flaw, reached the open internet, and broke into Hugging Face's production database. Those are two different decisions, and only one of them was supposed to happen.
Read
The Execution Layer Has Moved
If the next knowledge work role you fill is built around execution capacity — someone to do the research, draft the documents, produce the analysis — you're building for a version of the work that no longer describes how the work actually happens.
Read
Series
Foundations
The analytical and structural thinking that everything else builds on.
8 postsIn Production
What it takes to make AI work in production — the failure modes, misdiagnoses, and the questions worth asking.
5 postsCompound
How an AI operating system of specialized agents is designed, structured, and improved over time.
4 postsGround Truth
Plain observations on what AI is actually doing to knowledge work, organizations, and the people who lead them.
3 postsSubscribe to The Algorithm
Notes on building AI systems that actually work.