What has my attention right now.

Building EigenH AI

Most of my working time goes into EigenH AI. We are building a Patient Journey OS for independent healthcare practices, including specialty care: software for the entire patient journey, not just the front desk. The practical challenge is not simply making an agent speak. It is helping it act within a practice’s rules, preserve context, show its work, and return the right decisions to staff.

I’m spending a lot of time with the edges of that system: human handoffs, evaluation, memory, operating constraints, and the gap between an impressive demonstration and dependable daily use.

Writing more of the work down

I have years of notes from building products and leading teams. I’m turning the useful parts into essays, including ideas that are still changing. The goal is a record of how I think, not a stream of announcements.

Questions I’m carrying

• Where do AI agents remove real work instead of moving it?
• What should a system remember, and what should it forget?
• How do small teams keep judgment as their output increases?
• What does responsible automation look like in healthcare operations?

I update this page when the center of my attention changes.