Atatus
Cloud and AI Observability
Start new thread

AI SRE - AI that investigates production before you do.

by•
Atatus AI SRE investigates production incidents across logs, metrics, traces, infrastructure, and deployments, correlating evidence, identifying root causes, and recommending remediation before your team spends hours chasing signals. Now generally available for production teams.

Add a comment

Replies

Best
Maker
📌
Good to be here, Product Hunt. Sri from the Atatus team. Incident response still asks engineers to do too much manual work. An alert fires. Someone checks the metrics, searches through logs, opens a trace, checks recent deployments, compares service behavior, and starts forming a hypothesis. The telemetry is there, but connecting the evidence can take much longer than finding the signal itself. We built Atatus AI SRE to take on that investigation work. It brings together logs, metrics, traces, infrastructure signals, and deployment context to investigate an incident, identify the most likely root cause, show the evidence behind it, and recommend what to do next. The goal is not to replace the engineer. It is to give them a useful starting point instead of another dashboard to search. For teams handling production incidents regularly, that can mean less time spent collecting context and more time spent making the right decision. AI SRE is now generally available in Atatus. We'd genuinely like to hear how your team investigates production incidents today, and where the process still takes more time than it should.
Maker

Hey Product Hunt! Mohan from the Atatus team here 👋

One thing we kept seeing while building AI SRE: the hardest part of an incident often isn’t getting the data. It’s connecting everything fast enough.

Logs point one way. Metrics point another. A deployment happened 10 minutes ago. A pod restarted. An exception suddenly appeared. Engineers still have to connect those dots themselves.

That’s the problem we wanted AI SRE to tackle.

Instead of giving engineers another dashboard to search, it investigates across logs, metrics, traces, infrastructure, Kubernetes, and deployment context to build a failure path and surface the evidence behind the likely root cause.

We’re excited to finally put it in the hands of teams and learn where it helps, where it doesn’t, and what incident investigation still looks like in the real world.

Would love to hear how you currently approach a production incident.

Maker

Hey Product Hunt! Rishi from the Atatus team here 👋

One thing we cared about a lot while building AI SRE was making it useful during an actual production incident, not just generating a nice AI summary.

When something breaks, you usually have to jump between metrics, logs, traces, deployments, and infrastructure to figure out what actually happened.

AI SRE tries to connect those dots and show the evidence behind its findings, so you can quickly check whether the investigation makes sense before taking action.

For me, that trust part is really important. AI can help with the investigation, but the engineer should still be the one making the final call.

Would love to hear how others are handling this today.