Sentry introduces a visual agent, allowing developers to investigate any production problem in a simple language.

Application monitoring provider Sentry today announced the launch of Seer Agent, a new feature that enables developers to investigate and resolve production issues using natural language.

Seer Agent uses Sentry’s comprehensive telemetry to surface responses and communications that teams find difficult to find on their own, significantly reducing the time spent on configuring applications.

According to Sentry, developers can ask Seer Agent questions such as:

While AI coding tools have made coding faster, they have also created a new challenge: developers are increasingly cut off from the code itself. If it’s not someone on your team who puts in each line of code, bugs and incidents become difficult to investigate and diagnose. Seer Agent makes it easy to cut through the noise to understand exactly what’s going wrong, why, and how to fix it.

“When something breaks in production, you’re working on all the errors, spans, logs, metrics, and more, all at once. The volume of data alone makes it difficult to know where to start,” said Indragie Karunaratne, Sentry Senior Director of Engineering, said in the announcement. “The industrial agent questions all the sources, connects the relevant signals, and pinpoints what went wrong and where. Investigations that used to take hours now take minutes.”

Seer Agent is built on three key capabilities:

• Natural language questions: Ask any question about your app without needing to know where to look to Sentry.

• Connected context: Local relationships to all errors, scopes, logs, traces, and code context that a developer might never find by manually navigating.

• Agent investigation: Guides engineers through complex production problems by consulting evidence in real-time, uncovering what matters from Sentry’s big data.

“Most teams don’t struggle to know something’s broken. They struggle to know what to fix. Sentry has spent over a decade building production telemetry in response to that, and Seer is how we use it everywhere developers are already present – enabling more thorough root cause analysis, manual fix automation in code agents like Cursor and Claude Code, and open our data through the CLI, Dentry CEO, MCP announced Dentry. “On Slack, investigation becomes multiplayer. The dev team can merge the incident, redirect the Seer mid-step, and leave the thread as a record of how it was resolved. Seer Agent is another way to contact it.”

Seer Agent is now available on Slack, allowing users to initiate an investigation by sending a Seer message to any channel. It makes the experience multiplayer, by allowing anyone on the channel to ask, redirect mid-step, and add context the agent didn’t have before. Alternatively, channel participants can watch the team move from one event to another as a spectator to better understand the program.

More information is available here.