Canvas
AI agent · Data & analytics agent · Data & analytics by Honeycomb
Collaborate in an AI-guided investigation workspace to resolve production issues faster.
An AI-assisted copilot for developers and SREs that uses natural language to query telemetry data and explain production issues. It speeds up root cause analysis by generating visualizations and proactive hypotheses for SLO alerts.
Agent specs
The full spec card (context, pricing per token, modalities and more) on AntSpace. Sign up free to see the full spec card.
Category: AI Agents
- Channels
- In-app (Honeycomb platform)
- Integrations
- Honeycomb Observability Platform\nOpenTelemetry
- Deployment
- Cloud (SaaS)
- Availability
- Free, Pro, and Enterprise plans
- Interface
- Natural language and interactive workspace
- Data Foundation
- High-cardinality events and traces
- AI-assisted copilot for production debugging
- Natural language querying of telemetry data
- Plain-language explanations for SLO alerts
- Included in Free, Pro, and Enterprise plans
Overview
Canvas is Honeycomb's AI-assisted copilot designed for developers and SREs to investigate production issues through natural language interaction. It operates as a collaborative workspace where engineers can ask questions of their telemetry data, which Canvas then translates into queries and dynamic visualizations. The agent helps solve the "unknown unknowns" of complex distributed systems by surfacing anomalies and providing plain-language explanations for Service Level Objective (SLO) triggers. Canvas is included across all Honeycomb tiers, including the Free plan, to make expert-level observability accessible to every team member. By automatically triggering investigations in response to alerts, it allows engineers to arrive at a tested hypothesis rather than just a notification.
Investigation and Analysis
- Natural Language Querying: Users can query production telemetry data using natural language to explore high-cardinality events and traces.
- Automated Visualizations: The tool generates AI-driven visualizations to help engineers understand complex system behavior and patterns.
- Suggested Queries: Canvas suggests relevant queries based on the current context to speed up the investigation process.
- Interactive Data Exploration: Engineers can explore data interactively to pinpoint performance bottlenecks and root causes.
Alerting and Root Cause Analysis
- SLO Explanations: Canvas provides plain-language explanations of why a Service Level Objective (SLO) was triggered.
- Proactive Investigation: The tool can automatically trigger investigations in response to alerts, providing a tested hypothesis upon arrival.
- Anomaly Detection Integration: Canvas integrates with Honeycomb's anomaly detection to surface meaningful changes before they affect users.
Team Collaboration
- Shared Workspace: Teams can collaborate in a shared workspace to compare findings and share evidence during an incident.
- Evidence Sharing: Users can share Honeycomb links to specific investigations to facilitate team hand-offs and communication.
- Expert-Level Guidance: The tool allows even junior engineers to perform expert-level investigations through AI-guided workflows.
What it does
- Translates natural language questions into telemetry queries\nGenerates dynamic visualizations of production data\nExplains SLO violations in plain language\nTriggers proactive investigations for alerts\nSurfaces anomalies in high-cardinality data
Use cases
- SREs use Canvas to get plain-language explanations of why a specific SLO alert was triggered in production.
- Developers query high-cardinality telemetry using natural language to find the root cause of a latency spike.
- Engineering teams use the shared Canvas workspace to collaborate on incident evidence and compare findings in real time.
- Junior engineers leverage AI-guided investigations to perform expert-level debugging of complex distributed systems.
Specifications
- Channels
- In-app (Honeycomb platform)
- Integrations
- Honeycomb Observability Platform\nOpenTelemetry
- Data sources
- Honeycomb high-cardinality event store\nDistributed traces\nStructured logs\nTime series metrics
- Human control
- Engineers interact via natural language and review AI-generated hypotheses and visualizations in a collaborative workspace.
- Deployment
- Cloud (SaaS)
- Availability
- Free, Pro, and Enterprise plans
- Interface
- Natural language and interactive workspace
- Data Foundation
- High-cardinality events and traces
- Core Function
- AI-guided investigation and debugging
How it works
- Access the Workspace: Users access Canvas directly within the Honeycomb observability platform interface.
- Query in Natural Language: Engineers ask questions about system behavior or incidents using natural language.
- Analyze AI Evidence: Canvas generates queries and visualizations, suggesting next steps and explaining anomalies or SLO triggers.
- Collaborate and Resolve: Team members use the shared workspace to compare findings and confirm root causes based on the AI-provided evidence.
Pricing
- Free Plan
- Free forever
- Pro Plan
- Starting at $150 per month
- Enterprise Plan
- Custom pricing
FAQ
What is Honeycomb Canvas?
Canvas is Honeycomb's AI-assisted copilot that helps engineers investigate and resolve production issues by providing an AI-guided workspace for telemetry analysis.
Which plans include Canvas?
Canvas is included in all Honeycomb plans, including the Free, Pro, and Enterprise tiers.
How does Canvas use AI?
Canvas uses natural language processing to allow users to ask questions of their data, which it then translates into queries and visualizations.
Can Canvas help with SLO alerts?
Yes, Canvas can automatically trigger investigations when an SLO fires, so engineers arrive at a tested hypothesis instead of just an alarm.
Added by AntSpace from public sources (honeycomb.io).