Open-source AI agent monitoring platform. Latitude automatically detects all the ways your agents fail at scale, and gives your coding agent the tools to fix it.
Replies
Best
how far back does the weekly report actually go? is it a rolling 7 day window or does it compare week over week so you can see if usage is trending up?
@imogen_wallace We show you all your traces you ingested in the last 30 days for free, and it can be extended on paid plans. You can do all sorts of analysis and comparisons across any filters and windows.
If you're specifically asking about the Claude Code Wrapped weekly report, we do a full report on the current week's usage, and compare it with your last week's report!
Report
This looks like a great fit for teams running agents in production how does Latitude distinguish between an agent failing vs. just taking an unconventional but valid path to the goal?
Loved the idea and positioning! Congrats on the launch super excited to try it out
Report
Curious about the gives your coding agent the tools to fix it part does Latitude hand off a structured diagnosis to the coding agent or does it actually attempt the fix autonomously?
Report
How does this scale across thousands of concurrent agent runs? Is there a sampling approach or is every trace fully monitored?
Report
For teams already using something like LangSmith or Helicone, what's the core differentiator with Latitude is this more about detection depth or the auto-fix loop?
Replies
how far back does the weekly report actually go? is it a rolling 7 day window or does it compare week over week so you can see if usage is trending up?
Latitude
@imogen_wallace We show you all your traces you ingested in the last 30 days for free, and it can be extended on paid plans. You can do all sorts of analysis and comparisons across any filters and windows.
If you're specifically asking about the Claude Code Wrapped weekly report, we do a full report on the current week's usage, and compare it with your last week's report!
This looks like a great fit for teams running agents in production how does Latitude distinguish between an agent failing vs. just taking an unconventional but valid path to the goal?
EverTutor AI
Loved the idea and positioning! Congrats on the launch super excited to try it out
Curious about the gives your coding agent the tools to fix it part does Latitude hand off a structured diagnosis to the coding agent or does it actually attempt the fix autonomously?
How does this scale across thousands of concurrent agent runs? Is there a sampling approach or is every trace fully monitored?
For teams already using something like LangSmith or Helicone, what's the core differentiator with Latitude is this more about detection depth or the auto-fix loop?
Premast
What kinds of failure modes does it catch that traditional logging/observability tools typically miss?
How does Latitude handle false positives? I'd imagine flagging failures that are actually intentional edge-case behavior could get noisy at scale.
Is there a self hosted deployment option for teams with strict data residency requirements around agent logs and traces?