A physician-founder, an open-source library of AI agents for medtech, and the discipline that keeps them safe.
The combination of medical discipline, open-source thinking, and practical AI tools could genuinely change how early stage medtech gets built
Thanks Petar Exactly. The discipline is the moat. Eager to share more content soon
This is refreshingly grounded: narrow agents, source material, open prompts/code, and verification.
The next layer I’d want to see is consequence discipline: what the agent may do, what evidence it needs, when it escalates, and when it must stop.
Great comment Michael. Layering this is what grounds the entire thesis of ethically grounded agents. Agreed
Exactly. The agent layer only becomes trustworthy when the consequence layer is explicit: scope, evidence basis, escalation, and stop conditions.
The combination of medical discipline, open-source thinking, and practical AI tools could genuinely change how early stage medtech gets built
Thanks Petar Exactly. The discipline is the moat. Eager to share more content soon
This is refreshingly grounded: narrow agents, source material, open prompts/code, and verification.
The next layer I’d want to see is consequence discipline: what the agent may do, what evidence it needs, when it escalates, and when it must stop.
Great comment Michael. Layering this is what grounds the entire thesis of ethically grounded agents. Agreed
Exactly. The agent layer only becomes trustworthy when the consequence layer is explicit: scope, evidence basis, escalation, and stop conditions.