A Conversation with Arturo Oliver (ScienceLogic)
Stop focusing on reducing time for incident recovery (MTTR), AI-enabled observability tools can identify and prevent issues before they occur.
The Shift to Agentic Observability: From Insight to Autonomous Action
Observability isn’t just another name for monitoring, it’s an evolutionary step forward in collecting and using operational data in order to understand how IT systems are operating. Observability tools offer the promise of being able to automatically predict and even resolve issues before a failure occurs, provided that the process owners and stakeholders trust the tools to make system changes autonomously.
EAE Episode 32 – Show Notes
Summary
AI is reshaping the observability market, bringing greater ability to detect patterns and predict issues that may occur in the future. In this episode, Arturo Oliver (Head of Market Strategy, ScienceLogic) joins Dan Twing and Tom O'Rourke to discuss three major shifts: from reactive problem diagnosis to proactive prevention; from noise-filtering to AI-driven recommendations; and from scripted automation to agents that can take action independently. The episode tackles the harder, long-term challenge of connecting AI agents together across different tools and systems.
Key Points
- Observability is shifting from looking backwards to identify root causes towards a forward view that predicts problems before they occur
- AI agents are organized in layers with specialized roles, mirroring the complexity of real enterprise IT organizations
- Trust and transparency are the biggest barriers to adoption
- Traditional system performance telemetry is no longer enough, intent is also critical to understand why decisions were made and what outcome was expected.
- The "orchestrator of orchestrators" remains an unsolved challenge, aligning intent and policy across systems is still an area of exploration.
Recommendations for Automation Leaders
- Don't be a late adopter – look for quick wins with current agentic AI technology, don't wait for the technology to fully mature
- Prioritize prevention over recovery – focus on preventing incidents, not fighting fires
- Deliver intelligence to the right people at the right time – agentic systems should make your team more productive, not bury them in more data
- Trust and transparency are non-negotiable – AI insights and recommendations must be verifiable, traceable and auditable
- Focus on building institutional knowledge - capture and share your learnings as you apply agents and evolve new solutions
Links
Arturo Oliver: https://www.linkedin.com/in/arturooliver/
ScienceLogic: https://sciencelogic.com
EAE Podcast Home: https://em360tech.com/podcast-series/enterprise-automation-excellence
Feedback & Questions: eaepodcast@emausa.com