Agentic systems at scale, and the group that builds them.

The role

I lead AI agents and emerging products at System1, and I build agentic systems at scale. The work is both: product leadership for the group, and hands-on AI engineering on the systems themselves.

What makes it hard

Agentic systems at scale fail in the seams, so the job is to own the seams.

• Choosing the workflows where an agent beats a form or a search box.
• Making reasoning and tool calls dependable enough to run at scale.
• Building a group that can ship across design, engineering, product, growth, and marketing.
• Learning from what happens after launch, not only before it.

The group

Seven people across design, engineering, product, growth, and marketing, including two product managers. I set direction, write the specs, and stay close enough to the code to review it.

How I work there

The same method as everywhere else on this site: find the real problem, name what makes it hard, give each hard part an owner and a contract, and read the evidence after launch. Details of System1's products stay with System1.