FY26 Q3 Observability Roadmap Leadership Review

Attendees

@lmcandrew @abrandl @stejacks-gitlab @rnienaber @swiskow

On 30th October we met to review the current and future state of the Observabiliy Roadmap. We focussed on the recent team discussion and outcomes as documented in Observability Roadmap Sync - Summary Sep 2025 (#4352 - closed).

Immediate Priorities

  • Defer the ClickHouse evaluation project to focus on more foundational work that will improve our team effectiveness
  • Complete work in progress through Q4: Protocells observability support, SLA calculation framework, and Vector migration
  • Begin the standardized SDK and telemetry framework project, which Hercules will start picking up next week in collaboration with DevEx
  • Recognize that the SDK work is critical as it's a prerequisite for distributed tracing and provides the foundation for multiple other initiatives

Planning Requirements

  • Create a clear, detailed plan for the standardized SDK implementation with an explicit path showing how it enables distributed tracing, including realistic timelines that account for the foundational work required
  • Break down the Next-generation Service Catalog project into smaller, more manageable pieces that can deliver incremental value
  • Document how the service catalog enables AI-powered capabilities like agents and duo contributions
  • For Protocells observability work, document all components and future phases, recognizing this is an ongoing program rather than a single project

Distributed Tracing Roadmap

  • Develop a multi-quarter roadmap for Distributed Tracing that clearly communicates dependencies, costs (both tooling and potential fleet impact), and timeline
  • Have this roadmap ready before early next calendar year when questions about tracing will intensify as auth and other services come online
  • This documented plan will help us manage expectations, protect against scope creep, and potentially secure additional resources if leadership wants to accelerate the timeline

AI Strategy

  • Acknowledge there's a strong push for more AI capabilities across the organization
  • Identify 2-3 tangible, specific problems that AI can help solve rather than treating AI as a silver bullet for all our challenges
  • Ensure we fix fundamental issues (like standardized SDKs and Service Catalog) that cannot be solved by AI before pursuing AI-enabled solutions