Struct

Struct: The AI On-Call Agent for Automated Engineering Runbook Management

In the fast-paced world of software engineering, responding to system failures and bugs can consume hours of valuable developer time. Struct is an innovative AI on-call agent designed to automate your on-call runbook and streamline the incident response process. Backed by Y Combinator, Struct empowers engineering teams to proactively root-cause engineering alerts and bugs by cross-referencing logs, metrics, traces, and your entire codebase.

By the time an engineer opens their laptop, Struct has already completed hours of investigation, providing a clear path to resolution. Whether you are part of a startup or a large enterprise, Struct helps you stop triaging by hand and start resolving issues with data-driven precision.

What's Struct?

Struct is an intelligent platform that acts as an automated member of your on-call team. It is built to bridge the gap between receiving an alert and implementing a fix. By integrating context from your entire technical stack, Struct explores logs and metrics from leading observability platforms like Sentry and Datadog, as well as cloud providers and workflow tools like Slack, Linear, Asana, and GitHub.

Essentially, Struct serves as an AI investigation report generator. It automatically investigates every engineering alert as it occurs, replying with a detailed root cause, an impact analysis, and a suggested fix. This ensures that fast-moving teams can maintain high availability without the burnout of manual investigation.

Features of Struct

Struct offers a comprehensive suite of features designed to help engineering teams hit the ground running with automated on-call intelligence.

1. Automated Investigation and Root Cause Analysis

As soon as an alert is triggered, Struct begins its work. It automatically investigates the issue and provides:

• Root Cause Identification: Understand exactly what went wrong.
• Impact Analysis: See how the bug affects your users and systems.
• Suggested Fixes: Receive actionable advice to resolve the problem immediately.

2. On-Call Intelligence

Struct doesn't just react; it learns. Context from past issues, alerts, and investigations builds a repository of on-call intelligence that improves over time, making future resolutions even faster.

3. One-Click Fixes and Handoffs

Resolution is simplified through the Struct build agent. You can:

• Easily create Pull Requests (PRs) that always build cleanly.
• Hand off tasks to your own coding agent with full context included.

4. Deep-Dive Capabilities

For complex issues, Struct allows you to dive deeper directly within your workflow:

• Review collected evidence in Slack.
• Test alternative hypotheses.
• Explore incident timelines, commit histories, and log queries using AI investigation reports.

5. Secure by Design

Security is a core pillar of Struct. The platform is SOC2 Type II and HIPAA compliant. Your data is logically isolated, encrypted, and is never used for training models by Struct or its providers.

How to Use Struct

Getting started with Struct is designed to be a frictionless process that takes only minutes.

Step 1: Integrate Your Stack

Connect Struct to your entire stack to gather context. Struct supports all leading observability and alerting platforms including Sentry and Datadog, cloud log providers, and productivity tools like GitHub, Slack, and Linear.

Step 2: Automatic Investigation

Once integrated, Struct monitors your systems. When an alert occurs, Struct automatically investigates and posts a reply with the root cause and suggested fix. You can also @mention Struct in your communication tools to trigger an on-demand investigation for any specific issue.

Step 3: Review and Resolve

Use the evidence provided by Struct to verify the issue. You can review incident timelines and commit histories to understand the change that caused the alert. Finally, use the one-click fix feature to generate a PR or hand the task off to a developer.

Use Case Scenarios

• Proactive Root-Cause Analysis: A DevOps engineer receives a high-severity alert at 3 AM. Before they log in, Struct has already analyzed the logs and traced the error to a specific commit from the previous afternoon.
• Streamlined Incident Response: A fast-growing startup uses Struct to manage their on-call runbook, ensuring that every developer has the context needed to fix bugs without searching through fragmented logs.
• Efficient Handoffs: During a shift change, the outgoing engineer uses Struct to provide a complete incident timeline and impact analysis to the incoming lead, ensuring no information is lost.

Pricing Plans

Struct offers flexible pricing to support teams of all sizes.

• Free ($0/mo): Ideal for individuals. Includes 100 credits/mo, intelligent auto investigations, 1 user, and community support.
• Pro ($20/mo): Designed for small teams. Includes 220 credits/mo, Struct build agent, up to 5 users, and dedicated Slack support.
• Max ($200/mo): Our most popular plan for scaling teams. Includes 2,500 credits/mo, priority queue, higher rate limits, up to 25 users, and custom integrations.
• Enterprise: Tailored for large organizations. Includes unlimited users, RBAC & SSO/SAML, sidecar/on-prem support, and volume discounts.

Note: Annual discounts are available. 1 investigation typically consumes 15-30 credits. Credits can be topped up at any time and do not expire.

FAQ

Is Struct secure? Yes. Struct is fully SOC2 Type II and HIPAA compliant. All data is encrypted and logically isolated. Your data is never used for training purposes.

What does Struct integrate with? Struct integrates with all leading observability platforms (like Sentry and Datadog), cloud log providers, and work tools including Slack, Linear, Asana, and GitHub.

How do you install Struct? Struct is designed to be set up in minutes. You can start by integrating your stack and allowing Struct to begin building your on-call intelligence immediately.