OTel 101: The Standard Behind Modern Observability and AI

Most teams first look at OpenTelemetry to escape vendor lock-in. That's a real benefit, but it's not why OTel has become the standard behind modern observability and AI.

"AI" gets used to mean a lot of different things: AI-assisted coding tools, the LLM calls underneath your applications, autonomous agents that take action in production, and AI features built into your tools themselves. Each of these needs telemetry to be useful, and each fails in a different way when that telemetry is inconsistent or siloed.

OpenTelemetry is what makes any of them trustworthy: a consistent, vendor-neutral format for traces, metrics, and logs that any AI system can actually reason over.

In this session, we'll cover:

• What OpenTelemetry is and how it works, for those new to it
• Why "avoid lock-in" undersells the real value of a standard
• The different things people mean when they say "AI," and why each one depends on structured telemetry
• What this means for your data strategy over the next few years, not just your current tooling