Canvas Is Now GA: AI-Guided Observability for Modern Teams

When we introduced Canvas in beta, our goal was to reimagine how teams explore and collaborate around their observability data without requiring manual querying. Canvas has quickly become the AI-guided workspace that helps teams transform raw telemetry into meaningful, shared understanding faster than ever before. And today, we’re thrilled to announce that Canvas is now Generally Available (GA) for all Honeycomb users.

According to the 2025 State of AI-Assisted Software Development Report by DORA and Google Cloud, the industry has entered a new phase: “We’re still in the nascent stages of AI-assisted software development, a period of rapid technological and practical evolution.” The report highlights that AI’s true impact depends not just on adoption, but on deep integration with familiar workflows. Canvas embodies this by bringing AI directly to your observability data, ensuring your teams can ask questions, generate insights, and take action all in one place.

Built for the way modern teams work

Canvas isn’t about replacing engineers with automation. It’s about empowering them with a smarter, collaborative layer that accelerates discovery and learning. Teams can ask natural questions about their telemetry data as Canvas interprets intent and guides exploration. They can generate context-aware explanations of anomalies or spikes, drawing from traces, events, and metadata. They can collaborate live with teammates by embedding AI insights directly into shared analysis boards and connect seamlessly to existing Honeycomb datasets and workflows without switching contexts.

These capabilities align with the DORA AI Capabilities Model’s emphasis on AI-accessible internal data, healthy data ecosystems, and quality internal platforms, all proven to amplify AI’s benefits on organizational performance.

Since its beta launch, Canvas has seen adoption across hundreds of teams spanning every pricing tier, from fast-moving startups to major enterprise customers. Some of the most notable teams using Canvas today include Stripe, HelloFresh, Zeta Global, Duolingo, Robinhood, and many others. For many enterprise customers, Canvas has become a core part of their observability workflows and is the first tool they reach for.

Updates that you’ll see in GA

In addition to reliability investments that ensure Canvas stays on track, we’ve been working on developing stronger patterns for interactive and background Canvas use.

For interactive Canvas conversations, we are excited to introduce mention support where you can use an “@” symbol to bring up a search panel for querying across your Honeycomb datasets. When a dataset or column is mentioned in your conversation, Canvas automatically retrieves additional metadata—including information about recent sample values and ingest patterns—to improve the quality of results. We expect this functionality will make it easier for you to steer Canvas in a desired direction and to avoid dead ends in investigations.

We’ve also developed new support for background Canvas agents. With background agents, Honeycomb Intelligence can triage alerts and develop preliminary investigations before an engineer is paged. This speeds up investigations and allows human operators to focus on remediation if the root cause has already been identified. In the near future, you will also be able to configure Canvas to automatically begin investigating when anomalous patterns are detected.

What’s next

With Canvas now GA, Honeycomb customers can expect ongoing improvements in model guidance quality with better understanding of system context and query intent, deeper integration between Canvas and other Honeycomb surfaces, and richer collaboration and sharing capabilities across teams. As the State of AI-Assisted Software Development report notes, the future of engineering will belong to organizations that augment their systems to carry the gains of AI. Canvas is Honeycomb’s way of making that future tangible today.

Try Canvas today

Canvas is available to all Honeycomb users. Head into your workspace and click “Canvas” to start exploring your observability data with AI guidance. By collaborating together with Canvas, you can learn faster and build more resilient systems.