Question 1

Is Canvas available on all Honeycomb plans?

Accepted Answer

Yes. Canvas is generally available to all Honeycomb users. There is no separate add-on or special plan required to get started. Teams can access Canvas directly with the Honeycomb platform to explore and analyze their observability data with AI guidance at any time.

Question 2

How does Canvas work with existing Honeycomb datasets?

Accepted Answer

Canvas works directly with your existing Honeycomb datasets and does not require any new instrumentation or data modeling. It uses the same high-cardinality, event-based telemetry you already send to Honeycomb and applies AI guidance on top of it. Because Canvas operates on raw event data, it preserves full context during investigations and helps engineers explore what’s already there more efficiently, without duplicating data, transforming schemas, or introducing hidden abstractions.

Question 3

How does Canvas handle data access and permissions?

Accepted Answer

Canvas respects the same data access controls and permissions you already have in place in Honeycomb. This means users only have access to the datasets they are permitted to view, including the available fields and events in those datasets. There is no separate copy of your data created for Canvas, meaning it does not bypass any redaction, sampling, or data handling rules you’ve configured. Canvas is designed to augment investigation, not expand data access, adhering to today’s modern security and governance models.

Question 4

Does Canvas work with OpenTelemetry data?

Accepted Answer

Yes. Canvas works seamlessly with data ingested via OpenTelemetry, including traces, metrics, and logs. As long as your telemetry is available in Honeycomb, regardless of how it was collected, Canvas can help guide investigations and analysis.

Question 5

How does Honeycomb Canvas help engineering teams?

Accepted Answer

Honeycomb Canvas helps engineering teams diagnose issues faster by guiding investigations with AI-powered insights. Instead of starting from a blank query, engineers get guidance on what to explore next from insights like relevant filters to use, breakdowns, or relevant comparisons. All queries remain visible and inspectable, so teams can trust the results, understand why a suggestion was made, and debug with confidence during production incidents.

Question 6

What types of teams get the most value from Honeycomb Canvas?

Accepted Answer

Canvas is especially valuable for teams operating distributed, microservices-based systems with high-cardinality telemetry. It can also help teams with backend engineers, SREs, and platform engineering are responsible for reliability and performance. Additionally, teams working in fast-moving environments, such as cloud-native platforms or AI/LLM-powered applications, benefit from Canvas’ ability to guide open-ended investigations when failure modes are emergent and difficult to predefine.

Question 7

What query languages does Canvas support?

Accepted Answer

Canvas builds on Honeycomb’s native, event-based query model. Canvas works with the same queries and data structures already used throughout the Honeycomb platform. Instead of introducing another abstraction, Canvas helps engineers construct, refine, and iterate on existing honeycomb queries more efficiently.

Question 8

What's the difference between Canvas and traditional query builders?

Accepted Answer

Traditional query builders require engineers to manually decide which filters, groupings, and queries to try, often leading to trial-and-error exploration. Canvas shifts this workflow by introducing an AI-guided investigation workspace. Rather than replacing human judgment, it helps engineers focus on the most relevant signals by suggesting useful directions to explore based on patterns in the data.

Question 9

What's an AI-guided investigation workspace?

Accepted Answer

An AI-guided investigation workspace combines human intuition with machine-assisted analysis. In Canvas, AI helps surface patterns, correlations, and areas of interest in telemetry, while engineers remain in control of decisions and conclusions. The goal is to accelerate understanding and reduce cognitive load, not to hide system behavior or replace hands-on investigations.

Question 10

How do I debug production issues faster with AI?

Accepted Answer

To accelerate production debugging, use Canvas to systematically narrow the search space across high-cardinality telemetry. Instead of manually iterating through filters and groupings, Canvas applies AI to analyze how event distributions shift during an incident and surfaces the dimensions, fields, and comparisons most likely to explain the behavior change.

Because Canvas operates directly on raw Honeycomb events, engineers can pivot between traces, aggregates, and breakdowns without losing fidelity. Every suggested query is fully inspectable and editable, giving engineers the flexibility to validate hypotheses, understand causality, and converge on the root cause more quickly.

Question 11

Can AI write observability queries for me?

Accepted Answer

Canvas can assist in generating and refining observability queries based on your investigation context and intent. Engineers can use this assistance to quickly get started or explore unfamiliar data. All queries remain fully visible and editable, so the team can review, adjust, and iterate as needed.

Meet Canvas, where humans and agents collaborate

An AI co-pilot for your human team

Faster debugging with guided investigations

Shared workspaces for team analysis

Automatic incident response with investigations

Transform complexity into actionable insights

Ask questions in plain English

AI that shows its work

Surface relevant evidence

Bring your team up to speed

Make your agents experts

See it in action

Experience the power of Honeycomb

Related features

Honeycomb Intelligence

Honeycomb MCP server

Anomaly Detection

Want to know more?

Frequently asked questions