Documentation Index
Fetch the complete documentation index at: https://docs.struct.ai/llms.txt
Use this file to discover all available pages before exploring further.
Automatically root cause on-call issues
Struct is an AI support engineer that cross-references logs, metrics, traces, and your codebase to accurately root-cause engineering alerts and bugs — before you start debugging. Trigger an investigation by @mentioning Struct in Slack, assigning an issue in Linear or Asana, or letting Struct respond automatically to alerts in your connected channels.Struct completes its investigation before you open your laptop. Get started free — no credit card required.
How It Works
Integrate
Connect your observability stack, cloud logs, and codebase in minutes. Struct integrates with Sentry, Datadog, AWS, GitHub, Slack, and more.
Investigate
For every alert, Struct gathers logs, metrics, and traces — cross-references your codebase and recent commits — and delivers a structured root cause report.
Dive Deeper
Ask follow-up questions, explore alternative hypotheses, and query your production systems in natural language with Ask Struct.
Key Capabilities
Cross-Signal Correlation
Queries logs, metrics, traces, and your codebase in parallel — then connects the dots across systems.
On-Call Intelligence
Learns from past alerts and investigations to recognize patterns and improve accuracy over time.
Ask Struct
Ask questions in natural language across all your production systems. No more switching between dashboards.
Fix & Handoff
One-click PRs with the Build Agent, or seamless handoff to Claude Code and Cursor via MCP.
Get Started
Quickstart
Sign up free and run your first investigation in minutes
Connections
Connect your observability stack, cloud logs, and tools
Ask Struct
Query your production systems in natural language
Open in Claude Code / Cursor
Set up MCP server and CLI bridge for one-click handoff
Security
SOC2 Type II
Fully compliant with SOC2 Type II standards
HIPAA Compliant
Meets HIPAA requirements for protected data
- All data encrypted in transit and at rest
- Logically isolated per customer
- No customer data used for model training