Vladyslav Budichenko, Vocaly AI
AI agents can help with production incidents, but most demos skip the messy reality. This session covers using LLM-based agents and coding engines like Claude Code during real on-call work. You'll learn how agents can compress investigation time, improve MTTR, and help you understand what broke without drowning in alerts and dashboards. We'll look at what works: pulling context, correlating signals, suggesting fixes. And what doesn't: where agents give bad advice or waste your time.
Vladyslav Budichenko is a software engineer with 11 years of experience building distributed systems, AI applications, and production infrastructure at scale. An AWS- and GCP-certified professional architect, Vladyslav is also the founder of Vocaly AI, a platform to create voice AI agents for businesses. He is passionate about discovering new AI tools and integrating them into everyday workflows, especially for vibe coding and automations.
author = {Vladyslav Budichenko},
title = {{AI} Agents for Incident Investigation: The Good, The Bad, and The Ugly},
year = {2026},
address = {Seattle, WA},
publisher = {USENIX Association},
month = mar
}
