# Logfire traces and errors in Polylane investigations

Pydantic Logfire

> Published 2026-09-04 · Updated 2026-11-18 · Canonical: https://polylane.com/changelog/2026-09-04-logfire/

Teams on Pydantic Logfire can connect a project so Polylane watches its error rate, volume and latency, and investigations read its traces.

## What changed

Connect Pydantic Logfire and investigations can go from a failing route to the exact span that threw. Agents list the services that sent data in a time window, rank routes by request count, error rate and p50 and p95 latency, search spans and logs, and open one trace as a tree of parent and child spans. A read-only SQL query covers anything else, over a window of up to seven days.

Polylane watches each project's error rate, request volume and p95 latency like the other observability integrations, and names the service behind a rise in errors. See the [Logfire integration](https://polylane.com/integrations/logfire/).

## Why it matters

A Logfire project already holds the spans that explain a failure. Starting from the route that fails, an investigation can narrow to the one span that threw and read its exception, so the cause it reports points at a line of work, with the trace as evidence.

## How to use it

1. In Logfire, open your project's **Project settings > Security & access** and create a **Project** API key with **Read-only access** and **Manage project and org**.
2. From **Go to org settings > Security & access**, create an **Organization** key with the same access.
3. In Polylane, open **Settings > Integrations**, choose Logfire and paste the project key, then the organisation key. Connect each project separately; one organisation key works for all of them.

## Related

- [Logfire integration](https://polylane.com/integrations/logfire/)
- [Investigations](https://polylane.com/product/investigations/)



[All updates](https://polylane.com/changelog/)
