Everything We Talked About at O11yDay London 2026

O11yDay London focused on the operational challenges created by AI-assisted software development, arguing that code generation is becoming easier while review capacity, production reliability, agent behavior, token costs, and return on investment require greater attention. Speakers emphasized observability, experimentation, automated verification, and strong engineering foundations such as telemetry, testing, runbooks, and clear platform standards to manage agents responsibly. Examples from Honeycomb, Booking.com, Fin, Baseten, HelloFresh, Phorest, and others showed AI improving output and incident response while also creating risks including overloaded reviewers, incomplete investigations, hallucinated conclusions, hidden agent failures, noisy model behavior, and unpredictable spending. Several sessions advocated using human oversight according to system criticality, shifting low-risk review work to machines, measuring outcomes with production data and controlled experiments, and tracking agent-level traces, costs, retries, tool use, and accepted results. Across the event, presenters broadly agreed that AI systems are sociotechnical systems whose benefits depend less on promises of large productivity gains than on teams’ ability to instrument, evaluate, govern, and continuously improve them in production.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.