Oct 2, 2026 · 60 min
AI agents target enterprise work incumbents leave unfinished
Why AI Agents Can Beat the Incumbents
The episode examines whether AI-native companies can win enterprise software by automating the coordination, judgment, and edge cases beyond systems of record.
- 1AI-native companies can compete by handling complex procurement work that incumbent software largely leaves outside core systems.
- 2Multi-agent systems can coordinate sourcing, logistics, negotiation, and invoicing while reserving high-risk decisions for human experts.
- 3Durable vertical AI advantages may come from integrations, domain data, customization, and trust rather than foundation models alone.
Don't miss
Vlad walks through an airline-bolt procurement request, showing how agents handle routine sourcing and fulfillment while humans retain control of high-risk decisions.
The brief
Enterprise procurement spans legal, finance, engineering, logistics, and suppliers, exposing a gap between systems of record and the work required to complete a purchase.
Seema Amble and Vlad Kyle argue that incumbent chatbots cannot close that gap through retrieval alone; AI-native companies can build around process, policy, and principal agents.
Leo coordinates multiple agents across sourcing, negotiation, logistics, and invoicing, using shared context to move routine procurement from demand creation toward fulfillment.
An airline-bolt example shows the dividing line: agents can handle routine work autonomously, while experts remain involved in high-value negotiations and other risky decisions.
The broader bet is that vertical AI moats will come from integrations, workflows, memory, domain data, customization, and customer trust—not from a general model by itself.
Featuring
Listen to the full episode and explore every guest, topic, and moment on PodLume.