Math-First AI: How Kodamai Embeds Verification and Trust
from Amelia's Weekly Fish Fry Hosted by EEJournal.com
My podcast guest this week is Dr. Maha Achour, CEO and Founder of Kodamai. Maha and I chat about the shifting landscape of AI and how we can move beyond cloud and centralized data centers to a more distributed, agentic future. We discuss why governance must 'travel' with AI agents, what it means for trust to be architectural rather than procedural, and how Kodamai’s 'math-first' approach is helping engineers build verified, auditable AI systems that can thrive in the real world.
Embed this episode
Chapters
- 0:08 Episode intro & welcome
- 0:34 Topic preview & guest intro (Dr. Maha Ashour)
- 1:25 Why AI is becoming distributed
- 1:44 How we got here: human brain, math, data, agentic orchestration
- 4:35 Why governance must travel with agents (question)
- 4:56 Semiconductor analogy: embedding governance in compositional building blocks
- 6:58 Trust as architectural vs. procedural (question)
- 7:20 Building trust into AI: Calvin Grove, types & composable workflows
- 11:25 Future of distributed AI, embedded governance & hardware-software relationship (question)
- 11:37 Governance across tiers & autonomous vehicles example
- 14:45 Codamai’s agent platform focus (question)
- 15:07 Calvin Grove daemon: distributed typed microservices & hypervisor
- 17:10 Bridging industry and academia (question)
- 17:23 Industry–university collaboration & training the next generation
- 18:42 Codamai’s math‑first approach (question)
- 19:00 Types, category theory & neural‑symbolic augmentation
- 22:26 Off‑the‑cuff question: favorite meal
- 22:54 Closing: links, contact info & outro
- 24:29 Final sign‑off
No transcript yet.