#07Tuesday 2 December 2025

What’s It Doing That For? – Observability for AI in Production

—Jag Reehal ↗

As we all know, AI can be brilliant one minute and awful the next. Living in an unpredictable world of non-deterministic models can be… well, confusing. Especially as we move from prototypes to production, the question “what’s it doing now?” stops being a joke and starts becoming an operational concern.

This talk explores the emerging field of AI observability — the art (and occasional dark magic) of keeping LLM-powered systems reliable, explainable, and cost-effective once they’re out in the wild. We’ll look at practical approaches and tools for tracing behaviour across prompts, spotting drift, and catching those quiet regressions that sneak in after a model update.

From Code to Career: AI-Powered Skills Bridge

—Erdem Sahin ↗

This demo shows how to bridge skills gaps using AI automation. Coding skills as an example, it demonstrates building GitHub portfolios that showcase technical skills in interviews while delivering working features.

The session follows a 7-stage framework that transforms learning objectives into measurable outcomes. Attendees see live coding with AI APIs, automated processes, and quantifiable efficiency improvements.

Perfect for engineers transitioning to AI roles. The methodology applies across domains where manual processes can be intelligently automated.