AI observability: From reactive troubleshooting to proactive insights

How observability tools like AI Investigators, automated runbooks, and RAG improve workflows.

Tired of combing through endless logs or scrambling during incidents? In his lightning talk at All Things Open AI 2025, Wade Moore from Observe Inc. shares how AI-driven tools like automated runbooks, multi-agent orchestration, and retrieval-augmented generation (RAG) are reshaping observability. It’s a look at how the next wave of innovations will help engineers move from reactive troubleshooting to proactive, insight-driven operations.

The observability market is booming, projected to triple in size over the next decade. With massive data volumes flowing through platforms like Observe, the potential for AI-driven innovation is huge. Moore points to industry-wide adoption of open standards like open telemetry and Apache Iceberg, which are paving the way for smarter, community-driven tools that integrate seamlessly across teams and systems.

Read more: What is OpenTelemetry?

Central to this vision is the AI Investigator, a next-generation interface for exploring observability data. It enables real-time collaboration, step-by-step investigation tracking, and intuitive data exploration without deep institutional knowledge. Moore also highlights how AI agents are streamlining incident response by interpreting natural language instructions, triggering automated investigations, and eventually taking direct action across infrastructure and development environments.

Finally, Moore discusses how RAG connects AI models to real-time, context-rich sources like documentation, forums, and proprietary query languages. While live system data integration is still in development, it promises the ability to query complex, streaming datasets and receive actionable insights in seconds once fully realized.

Read more: Observability is confusing, here’s how to learn it

Key takeaways

•  The observability market is primed for AI innovation, fueled by vast data volumes and open standards like open telemetry and Apache Iceberg.
•  AI Investigator tools will make data exploration collaborative, intuitive, and traceable, replacing siloed, manual workflows.
•  Multi-agent orchestration and RAG promise faster incident resolution by combining live system data with contextual knowledge.

Conclusion

The future of observability is moving toward proactive, automated, and deeply integrated AI solutions. As Moore’s talk shows, the groundwork is already being laid and now it’s just a matter of how quickly these tools can evolve to meet the scale and complexity of modern systems.

More from We Love Open Source

• What is OpenTelemetry?
• Observability is confusing, here’s how to learn it
• Are we coding through a revolution or an evolution?
• How to build a multiagent RAG system with Granite