One operational timeline.
Put code changes, deployments, and production signals in the same chronological record.
Building in the open Follow along See what changed
See what changed
before production broke.
Code changes, deployments, alerts, and incidents.
One timeline to investigate what happened. Evidence to explain why.
Release & incident intelligence for small engineering teams.
OpsReplay / Incident workspacePRODUCT PREVIEW
SEV 2Investigating
Checkout API latency spike
checkout-apiproductionToday, 09:56 UTC
One record. From the first change to the final review.Illustrative product preview · Early development
The context you need.
Already in sequence.
Code changes Deployments Production signals One timeline
CONTEXT, WITHOUT THE TAB HOPPING
The incident is one story.
Your tools tell it in pieces.
OpsReplay is being built to connect those pieces, so your investigation starts with the sequence of events.
The change before the signal.
See the commits and deployments that preceded a failure. Know where to start looking.
Deployment
09:475 min
First errors
09:52Evidence you can follow.
Keep the investigation tied to source events. Follow a reference back to the record behind it.
event_0248source record
Postmortems with receipts.
Draft the report from collected evidence, with traceable references and room for human judgment.
PLANNEDGrounded drafts, reviewed by you.
Built for the smaller team.
A focused release and incident workflow, without a sprawling enterprise incident platform.
Less overhead. More context.
FROM CHANGE TO UNDERSTANDING
A continuous record.
A clearer investigation.
The workflow we’re building: capture the events first, then connect the incident to the evidence.
- 01
Connect your tools
Start with GitHub. Deployment and monitoring sources are next.
- 02
Build the record
Capture timestamped events as your software changes.
- 03
Follow the sequence
See what happened before, during, and after an incident.
- 04
Write from evidence
Turn the recorded context into a referenced draft for review.
EVIDENCE FIRST. AI SECOND. PLANNED
Every claim should
have a way back.
The planned postmortem assistant draws exclusively from the collected incident evidence. A referenced draft to review, with uncertainty kept visible.
INCIDENT REPORT
Checkout API latency spike
Summary
Errors and elevated latency were observed in the Checkout API. An incident was opened at 09:56 UTC.
Probable contributing change
Deployment 7ac04e2 occurred five minutes before error rates began increasing.
!
Timing suggests a change to investigate.
Root cause is not yet confirmed.
Next steps
Review the deployment diff and validate the connection-pool behavior before attributing a cause.
SUPPORTING EVIDENCE3 events
Deployment01
Production deployment completed
7ac04e2checkout-api
Evidence
Monitoring02
Error rate begins increasing
5xx responses above baseline
Evidence
Incident03
Checkout API latency spike
INC-042Severity 2
Evidence
Reference 1 selected · deployment at 09:47
Select a numbered reference to see its source.Illustrative draft · No live AI calls
MEET YOUR TOOLS WHERE THEY ARE
Your stack.
A shared context.
GitHub is the starting point. More event sources will bring the rest of the production story into view.
Follow integration developmentGitHub IN DEVELOPMENT
App connection, verified webhooks, and push events in the repository timeline.
Deployment systems
Releases, rollouts, and rollbacks
UPCOMINGMonitoring & alerts
Errors and production signals
UPCOMINGCloud infrastructure
Changes to the systems underneath
UPCOMINGIncident systems
The response, alongside the cause
UPCOMINGBuilt in the open.
From the event pipeline up.
FastAPI / PostgreSQL / Docker / SQS-compatible queues
Explore the sourcePUT THE EVENTS IN ORDER
Your incident timeline shouldn’t
start after the incident.
Follow OpsReplay as we build a clearer path
from production changes to incident understanding.