Agents are where microservices were in 2015: Navan engineering insights
AI Engineer · youtube · 2026-08-30
Navan engineers Roberto Milev and Uday Kanagala discuss the state of AI agents, comparing the current hype to the microservices era of 2015—advising against complex agentic loops when a single loop suffices. They advocate for a "master agent + skills" architecture where skills act as pluggable, testable context units. To address logging challenges with verbose reasoning, they use hooks to capture goals, reasoning, and confidence scores at every tool call. Testing focuses on scoring trajectories rather than asserting outputs. They also highlight unresolved issues around cost, authorization ambiguity (e.g., who bought the flight?), and the need for emerging standards.
More from coding & agent
- AI tools eroding code review: team reads only 5% of PRs — Upset-Day9099 · 2026-08-30
- Tesana Launches Loop Mode for Self-Testing 3D Game Generation — Aiden_Tech_Ai · 2026-08-30
- Open Source quackd: Control Bipedal Robot via Natural Language — DynamicWebPaige · 2026-08-30
- Omarchy spins up agent to diagnose crashes via core dump — steren · 2026-08-30
- Publishing to the Claude Connector Directory: 60 signups in 2 days, 3 sales, then the discoverability cliff — kkomelin · 2026-08-30
- Adobe's Agentic Sites assembles hyper-personalized pages in ~1.1s, no frontier model needed — AI Engineer · 2026-08-30