Split with NeuBird, the AI SRE: the flag that changed it
NeuBird reads Split feature flags, targeting rules, segments and change requests, then checks whether a flag change lines up with the incident.
< 5 min
Root cause, at 94% accuracy
15+
Sources queried in parallel per investigation
In place
No bulk telemetry stored
< 1 hour
To deploy on the stack you already run
Core Capabilities
From signals to solutions
Resolve
A flag flip is a change too
When an alert fires with no recent deploy, NeuBird lists the Split feature flags in the affected environment, reads their targeting rules and recent change requests, and lines the timing up against your metrics, logs and traces. It tells you whether a rollout or a targeting change started the incident.
- Checks flag changes in the right environment against the incident window
- Names the flag, treatment and segment behind the change
- Root cause in under 5 minutes at 94% accuracy, posted where your team works
Prevent
Watch the rollout, not just the deploy
NeuBird watches the services behind a flag as its rollout widens and flags degradation 30 to 60 minutes before it would page, while turning the flag back is still the simplest fix.
- Flags services degrading as a feature flag reaches more traffic
- Surfaces flags rolled out to production with no alert on the services they touch
- Recommends SLOs for the features you release behind flags
Operate
Flag history in every investigation
Between incidents, NeuBird looks across past investigations for the flags and segments that keep appearing in root causes, so your team can tighten the release process where it matters.
- Identifies flags that recur in incident root causes
- Recommends turning a flag back, behind configurable human approval gates
- Remembers every verified fix, with no bulk telemetry stored
Better Together
Split + NeuBird
| Capability | Split | NeuBird |
|---|---|---|
| Feature flags and progressive rollouts | ✓ | Reads them in place |
| Targeting rules, segments and change requests | ✓ | Uses them as investigation evidence |
| Correlation with metrics, logs, traces and deploys | Feature impact data | 15+ sources in parallel |
| Root cause with cited evidence | Not in scope | Under 5 minutes, 94% accuracy |
| Staged fix or rollback | Kill switch | Behind configurable human approval gates |
| Memory of past fixes | Change history | Verified fixes, no bulk telemetry stored |
Ecosystem
Works across your entire stack
Split is one piece of the picture. NeuBird queries it in place and correlates it with every other connected tool, and every agent in the center inherits the connection.
Delivery
- GitHub
- GitLab
- Argo Workflows
Observability
- Datadog
- New Relic
- Grafana
Incident management
- PagerDuty
- Opsgenie
- ServiceNow
Collaboration
- Slack
- Microsoft Teams
FAQ
Common questions
How does NeuBird connect to Split?
You create a Split Admin API key and enter it with your Split API URL in NeuBird. NeuBird queries Split at investigation time and stores no bulk telemetry.
What Split data does NeuBird use?
Environments, feature flags and their definitions, including targeting rules and treatments, segments and change requests.
Does this work now that Split is part of Harness?
Yes. Harness acquired Split in 2024 and offers it as Harness Feature Management and Experimentation. NeuBird connects with a Split Admin API key and your Split API URL.
Is NeuBird a replacement for Split?
No. Split keeps managing your feature flags and experiments. NeuBird is the AI SRE that checks flag changes during an investigation, alongside every other tool you connect.
Get Started
Connect Split to the center.
Split gives you the data. Connect it to NeuBird once, and every agent that touches production inherits it, including the ones you build yourself.
