How Expedia Rebuilt Its Platform for AI Agents Across 30k Repos
Thanigai Vellore, VP of Engineering at Expedia, on rebuilding a developer platform for 30,000 repos when AI agents broke every internal tool built for humans.
This is 1 IDEA.
Every episode, 1 engineer explains 1 great idea. That's the show. No buzzwords. No fluff. No marketing. Engineers only.
Subscribe to the show
Past Episodes
Why AI Made Us Redesign Our Distributed System
Sedai's VP of Engineering on re-architecting a distributed system to expect failure, moving fault tolerance to a central orchestrator so it recovers on its own.
Why Your AI Agents Keep Failing in Production
On this episode of 1 IDEA, Jason Li (CTO @ Laurel) breaks down how to separate deterministic and probabilistic systems, what that means for incident response, and why AI code generation just made code review the most important bottleneck in your pipeline.
Why AI Isn't Actually Making Eng Teams Faster
In this episode of 1 IDEA, Suresh Mathew sits down with Sashanka Vishnuvajhala (SVP of Engineering @ Fanatics) to break down why and how he rebuilt the entire pipeline from spec to production.
Is Your CI/CD Pipeline Actually Working?
Satya Singeetham (CTO @ New Horizon Bank) is a world-leading expert on CI/CD Pipelines. In this episode of 1 IDEA, he explains how to build a world-class CI/CD setup in 2026.
Validate Your ML Training Data Before Production Does It for You
On this episode of 1 IDEA, Eugene Fratkin (VP of Engineering @ Salesforce) breaks down why ML training data fails in production and what to validate before you ship.
How 30 Years of Tech Prepared This CTO For AI
On this episode of 1 IDEA, Suresh Mathew sits down with Kevin Trilli (CPTO @ FIS Amount) to unpack his playbook for getting ahead of technology transitions, like the AI era.
How We Solved the GPU Problem for Kubernetes
We solved GPU optimization. In this episode of 1 IDEA, Suresh Mathew sits down with Pooja Malik, Distinguished Engineer at Sedai, to talk through how Sedai's engineering team built GPU optimization from scratch, why the standard metrics fail, and what's still unsolved.
The Startup Within A Startup Playbook
John Wang (Co-Founder @ Assembled) did what most companies talk about but rarely pull off: he spun up a four-person AI team inside an existing company, ran it like a YC startup, and hit millions in ARR in 18 months. But the board was angry, the org nearly split, and the sales team couldn't sell the product.
How To Prepare For The Next AWS Outage
In this episode of 1 IDEA, Suresh Mathew sits down with Preeti Somal (SVP of Engineering @ Temporal) to unpack how her team prepared for a failure they knew was coming and how other engineering leaders can do the same.
The 23-Minute Playbook for On-Prem Products
Gil Feig (CTO @ Merge) took on what everyone warned him not to: on-premises deployment. 18 weeks to first customer, and a support model that could sink the company.
How 4 Engineers Shipped AWS's Fastest Product Launch
Ashish Jha (now Director of Engineering @ Drata) built AWS Audit Manager — one of AWS's fastest-growing services ever — launching in 15 months with 1,000 customers turning it on in the first 24 hours.
Resources
The Bottleneck Test | Vinay Pereti
Vinay Perneti (VP of Eng at Augment) shares the Bottleneck Test: a framework for leaders to empower their teams when they become the blocker.
Minimum Lovable Product | Mohit Aron
Mohit Aron built 2 companies to $10B+ (Nutanix & Cohesity) by rejecting Silicon Valley's MVP obsession. He reveals why a "Minimum Lovable Product" beats an MVP every time.
Join our Community
SubscribeWatch on YouTube
1 IDEA
Subscribe to 1 IDEA
Sign up for notifications when episodes drop.
Recent resources
The Man Who Wrote the Book on Micro Frontends
Luca Mezzalira, who literally wrote the book on micro frontends, explains everything you need to know
Sed Demo Video
Watch a few of the coolest things you can do with Sed, your cloud & AI assistant, in this short demo
Sed Solution Sheet
Learn more about Sed: The cloud & AI assistant from Sedai.
Meet Sed: Your Cloud & AI Assistant
Sed is an AI assistant that answers questions & makes changes across your entire stack
How Expedia Rebuilt Its Platform for AI Agents Across 30k Repos
Thanigai Vellore, VP of Engineering at Expedia, on rebuilding a developer platform for 30,000 repos when AI agents broke every internal tool built for humans.
How Palo Alto Networks Takes Control of Its High-Stakes Cloud
Learn how Palo Alto Networks dramatically reduced its cloud costs with Sedai
Why Your AI Agents Keep Failing in Production
On this episode of 1 IDEA, Jason Li (CTO @ Laurel) breaks down how to separate deterministic and probabilistic systems, what that means for incident response, and why AI code generation just made code review the most important bottleneck in your pipeline.
Why AI Made Us Redesign Our Distributed System
Sedai's VP of Engineering on re-architecting a distributed system to expect failure, moving fault tolerance to a central orchestrator so it recovers on its own.
You're Debugging AI Failures the Wrong Way
On this episode of 1 IDEA, Jason Li (CTO @ Laurel) shares his rule on what actually breaks AI in production and what to do about it.