AI Chat: ChatGPT & AI News, Artificial Intelligence, OpenAI, Machine Learning Agentic AI Engineers with Spiros Xanthos, Resolve AI CEO
13 snips
Aug 28, 2025 Spiros Xanthos, CEO of Resolve AI, is on a mission to revolutionize the engineering landscape with agentic AI systems. He discusses how these AI tools automate the detection and resolution of production incidents, drastically improving productivity. Spiros highlights their proactive capabilities in incident prevention, transforming traditional roles for engineers. The conversation also touches on ensuring AI reliability, reducing critical errors, and the future of creative collaboration between humans and machines in software development.
AI Snips
Chapters
Transcript
Episode notes
From Observability To Agentic Automation
- Spiros and his co-founder built observability tools and OpenTelemetry before Resolve AI existed.
- They spent more time firefighting production systems than building, which inspired agentic automation.
Effectiveness Grows Over Time
- Agentic systems can handle many repeat incidents and generalize to novel problems by reasoning from symptoms.
- The key metric is improvement over time (day 30, 60, 90) as agents learn from the environment.
Verify Claims And Measure Real Performance
- Build strong evaluation sets and measure where agents succeed or fail before deploying widely.
- Require agents to verify claims with citations and present chains-of-thought for quick human validation.