Notes on AI systems design, digital transformation, trustworthy AI, and what 25 years of large-scale engineering can teach us about getting this generation of technology right.
Most enterprise AI strategies are built on a quiet assumption: that intelligence is the hard part. After two and a half decades engineering systems for national laboratories, I'm convinced it isn't. The hard part is everything around the model.
A working framework for building AI systems that survive contact with auditors, scientists, and skeptical end-users — drawn from deployments in research and federal environments.
Notes from leading enterprise transformation across research, government, and commercial environments. The common patterns of failure are remarkably consistent.
New essays publish roughly monthly. Follow on LinkedIn for first access — discussion happens there.
Follow on LinkedIn →