Why most institutional AI pilots stall at month four
A pattern we see across hospitals and universities — and the four interventions that unstick it.
Coming soon
Writing on adoption, evaluation, governance, and the unglamorous work of moving AI from notebook to production at institutions that don't get to fail loudly.
Eight pillars, one honest read: where the AI for All draft is strongest, where the delivery gap is, and what we'd build first.
Published →
A pattern we see across hospitals and universities — and the four interventions that unstick it.
Coming soon
If you can't measure your AI system's quality on your own data, you don't have a system. You have a vendor demo.
Coming soon
A governance pattern small enough to use, structured enough to defend to a board or a regulator.
Coming soon
What changes when you treat your retrieval layer as a first-class part of the user experience.
Coming soon
New essays in the pipeline. In the meantime, read 20+ years of writing at hammad.blog.