Practical tools to simplify GenAI observability and monitoring

Learn why developers need full-stack observability for AI-driven apps.

Adrian Cole, principal engineer at Elastic, sat down with the All Things Open team to share how open telemetry and generative AI are reshaping observability and developer productivity.

Read more: What is OpenTelemetry?

Adrian explains that open telemetry provides a shared set of specifications, SDKs, and infrastructure to help developers monitor and understand their applications. Going beyond traditional monitoring, it includes tracing and profiling, making it easier to capture the full picture of how apps behave in real time.

He highlights how generative AI intersects with observability. Unlike predictable database requests, GenAI outputs are harder to track and evaluate. Developers need complete records of inputs and outputs to measure quality, which requires new observability approaches. Adrian also shares how AI assistants are making monitoring tools like Kibana more accessible, letting developers query data with natural language and automate configurations.

When asked about AI’s impact on development work, Adrian sees AI as an amplifier rather than a replacement. He points to “vibe coding” for quick prototypes and research tasks where AI can save hours, while still relying on developers for higher-value engineering. He also shares two favorite tools: Goose, a generative AI assistant from Block for software engineering tasks, and Hotel TUI, a terminal-based interface for open telemetry that speeds up log and metric analysis.

Key takeaways

• Think beyond monitoring: Use open telemetry to capture tracing, profiling, and full application behavior.
• Pair GenAI with observability: Track complete inputs and outputs to evaluate AI-driven responses and ensure quality.
• Start small with AI: Paste open source code into an AI tool to get explanations and learn faster.

Conclusion

Adrian’s insights show how combining open telemetry and generative AI can boost developer productivity without replacing human expertise. By embracing these tools and starting small, developers can build smarter, more observable systems while keeping control of their workflow.

More from We Love Open Source

• What is OpenTelemetry?
• 6 limitations of AI code assistants
• Your 2025 guide to AI, no-code, and developer-led software
• How I use AI agents to automate my workflow and save hours
• Why Kubernetes is essential for AI and open source projects in 2025