Scaling Delivery Infrastructure Without Scaling Your Observability Bill: Lessons from Jitsu's CTO
Grepr
44:12
A live conversation between Jad Naous (Grepr) and Evan Robinson (Jitsu)
Jitsu delivers millions of packages a year for leading brands around the world. Behind every delivery is a real-time tech stack doing route optimization, driver dispatch, geofencing, image-based proof of delivery, and customer communication. Every one of those systems generates data. A lot of it.
Evan Robinson, Jitsu's CTO, has spent his career building platforms at this kind of scale. He co-founded Shipwire (acquired by Ingram Micro), ran a platform engineering team at Yahoo, and now oversees the AI-powered infrastructure behind Jitsu's 99%+ on-time delivery rate across 122+ million people worth of U.S. metro coverage.
In this fireside chat, Grepr founder Jad Naous sits down with Evan to talk about the engineering reality of running AI-intensive operations at national scale, and what happens to your observability costs when you do.
What they'll cover
- What Jitsu's infrastructure actually looks like behind the scenes, and how it changed as they scaled from a few metros to dozens of cities
- Where observability spending started creeping up, and when it became a leadership-level conversation
- How Evan's team decides what data is worth paying to keep vs. what's noise they can afford to compress or drop
- Real incidents where having the right telemetry made the difference, and times where too much data slowed the team down
- What AI in production operations looks like at Jitsu today, from multimodal delivery analysis to self-healing systems
- Practical advice for engineering leaders trying to get observability costs under control without flying blind
Who this is for
Platform engineers, SREs, DevOps leads, and engineering managers who are responsible for infrastructure spend and observability tooling. Especially useful if your team is running Datadog, Splunk, New Relic, or similar platforms and you've started asking hard questions about the bill.
Format
35-45 minutes. No slides, no pitch deck. A live conversation followed by open Q&A where you can ask Evan and Jad anything. Streamed simultaneously on LinkedIn Live, YouTube Live, and X.
Can't make it live? Register anyway and we'll send you the recording.
Speakers
Jad Naous
CEO, Founder @ Grepr
Jad Naous is the founder and CEO of Grepr, where he's tackling observability cost optimization after more than a decade building and investing in data infrastructure. Before Grepr, he shipped products at AppDynamics, invested in infrastructure deals at Andreessen Horowitz, and scaled engineering at Imply, the company behind Apache Druid.
Evan Robinson
CTO @ Jitsu
Evan Robinson is the CTO at Jitsu, where he oversees the AI-powered technology platform behind last-mile delivery for leading brands. He previously co-founded Shipwire, building it from scratch into a fulfillment platform that competed with Amazon before selling it to Ingram Micro in 2013.
Scaling Delivery Infrastructure Without Scaling Your Observability Bill: Lessons from Jitsu's CTO
44:12