Prerequisites
- A GitHub account with repository access
- A log provider (AWS CloudWatch, Azure Monitor, GCP Logging, Datadog, Loki, SumoLogic, etc.)
- An Edgedive account (sign up here)
Setup takes approximately 15 minutes. White-glove onboarding is available—reach out to team@edgedive.com. Step 1: Connect GitHub
Required. GitHub is how Edgedive accesses your codebase for investigations and creates pull requests.
Go to Connections
In Edgedive, navigate to Connections
Connect GitHub
Click Connect next to GitHub and install the Edgedive GitHub App
Select Repositories
Choose which repositories Edgedive should have access to
Start with a single repository to test the workflow before expanding access.
Step 2: Connect a Trigger Source
Connect Slack or a ticketing system so Edgedive knows when to investigate. You can @tag Edgedive, assign an issue to it, or have it respond automatically to alerts.
Connect Slack
Go to Connections → click Connect next to Slack
Authorize
Authorize the Edgedive app in your workspace
Invite to Channels
Invite @Edgedive to channels where alerts are posted or issues are discussed
@Edgedive investigate high latency on the /api/checkout endpoint
Connect Linear
Go to Connections → click Connect next to Linear
Authorize
Authorize and select teams to monitor
Step 3: Connect a Log Provider
Give Edgedive access to your production logs for deeper investigations. This can be a cloud log provider or an observability platform.
Cloud Logs
Observability Platforms
| Provider | Setup |
|---|
| AWS CloudWatch | Go to Agent → Logging Access Providers → select AWS CloudWatch |
| Azure Monitor | Go to Agent → Logging Access Providers → select Azure Monitor |
| Google Cloud Logging | Go to Agent → Logging Access Providers → select Google Cloud Logging |
| Platform | Setup |
|---|
| Datadog | Go to Connections → Datadog |
| Loki | Go to Connections → Loki |
| SumoLogic | Go to Connections → SumoLogic |
| Sentry | Go to Connections → Sentry |
Cloud log access enables Edgedive to query production logs in real time, dramatically improving root cause accuracy.
Step 4: Trigger Your First Investigation
From Slack
From an Issue Tracker
Automatic (from alerts)
In any channel where Edgedive is present:@Edgedive investigate 500 errors spiking in the payments service
Assign an issue to Edgedive or @tag it in a comment. Edgedive picks it up and starts investigating automatically.
Once connected, Edgedive can automatically investigate incoming alerts from your connected sources. Wait for your next alert, or trigger a test alert.
Step 5: Review the Investigation Report
Edgedive delivers a structured report with:
- Root cause analysis — What went wrong and why
- Impact analysis — What’s affected and the severity
- Incident timeline — Key events leading up to the issue
- Suggested remediation — How to fix it
From here, you can:
| Action | How |
|---|
| Ask follow-up questions | Reply in the Slack thread or issue comment |
| Create a PR | Click Approve to let Edgedive implement the fix |
| Investigate further | Ask Edgedive to dig deeper into specific signals |
That’s it! Edgedive will now investigate alerts from your connected sources.
What’s Next?
