Sentry

Developers usually learn that code is broken when a customer complains, and Sentry is software designed to tell them first. It is an application monitoring platform. When code fails in production, whether through an exception, a crash or a sluggish request, Sentry records the event with surrounding context so a team can tell which users, services and releases were hit. Its error tracking is open source.

Starting small

The company stresses a light setup: drop in the SDK in about five lines, with no agents to install and no performance surprises. Because the SDKs cover many languages and frameworks, including web and mobile, errors anywhere in the stack end up in the same place.

Context in one trace

Sentry's argument is that errors, logs, session replays, spans, profiles and metrics are most useful when connected by the same trace. Instead of a stack trace alone, a developer moves from the issue to the context around it and then to a fix. Source code, error filters and stack locals add detail that logs and user reports tend to miss.

Automated root cause and AI features

The product includes Seer, a debugging agent that also reads code changes, which looks for the root cause of an issue and can flag problems before code is merged. Agent tracing helps teams that build AI features find unexpected output and bad tool calls and see spend and performance across their agents. An MCP server lets developers pull Sentry data into an LLM-based workflow to fix production issues.

Fitting existing workflow

Sentry connects with GitHub, Slack, Jira and Linear, so alerts and fixes flow into tools already in use. Earlier posts on its site discuss dynamic sampling, which was made available to new customers, and mobile development with SwiftUI and Jetpack Compose.

Who uses it

The company describes itself as developer-first, and says millions of developers rely on it.

Sentry is an application monitoring platform built for developers. It captures errors, crashes and slow requests in production code and attaches context, so teams can see which users, services and releases were affected and fix problems sooner. Its error tracking is open source.

Features

• Error tracking: groups crashes and exceptions with stack traces and surrounding detail.
• Performance monitoring: follows requests end to end to find slow calls and queries.
• Releases and issue owners: shows which errors a release fixed or introduced, and who owns them.
• Seer debugging agent: an AI tool that analyses issues and proposes fixes.
• Session replays, logs and profiles: linked to the same trace as errors.
• Dashboards and Discover: visual views and a query builder for event data.

Customers and markets

Software engineering teams, and also support and product staff who need to see application health. It works with many languages, and integrates with GitHub, Slack, Jira and Linear.

Key facts

• Setup: an SDK is added to the code, with no agents to install