Jad Naous is founder and CEO of Grepr, an observability and operational-reliability company. He previously led architecture at AppDynamics and now focuses on the operational work surrounding AI-generated software.[1][2][3]
Start with Why AI Can Write Code But Can’t Keep It Running (Yet) →
X Career
Grepr’s founders came from observability companies, including AppDynamics, Imply and New Relic. Its company biography identifies Naous as CEO and credits his earlier architecture leadership at AppDynamics. Grepr works on reducing the volume and cost of operational telemetry while retaining useful signals for engineers. The company’s broader focus is making systems easier to operate as AI accelerates the production of code.[1][3]
Ideas
In his a16z conversation with Martin Casado, Naous argues that producing code does not settle how the resulting system behaves under changing workloads and failures. The discussion considers troubleshooting through change detection, hypotheses and experiments. It also examines how coding agents alter a small engineering team’s workflow. The reported speedups are experiences from his team, not controlled evidence that every software project will improve by the same amount.[2]
Tools, writing & talks
Writing, research & talks
Start here
An interview about coding agents, observability and the harder task of operating production software.
Start with the discussion of why production behavior is harder to reproduce than a code-writing task. Experience running an application helps connect the examples to incidents and observability.
Interview guest[2]Original source