Find the Root. Fix the Cause.
AI-Powered Root Cause Analysis for SRE & Performance Engineering
Stop staring at charts. Let AI explain what went wrong and how to fix it.
Get Started:
pipx install heimr-ai
Supports Your Stack:
🚀 Coming soon: NeoLoad, Grafana, Kibana, OpenTelemetry, and more...
Explainable Analysis
No black boxes. See exactly why something was flagged with statistical anomaly detection using Z-scores and IQR.
Multi-Signal Intelligence
Correlates load tests, system metrics, application logs, and distributed traces for comprehensive root cause analysis.
LLM-Powered Insights
Get root cause analysis in plain English. Works with OpenAI, Anthropic Claude, or run locally with Llama/Qwen.
Privacy-First
Run 100% locally with Ollama. Your performance data never leaves your infrastructure.
Performance Comparison
Compare runs, detect regressions, and track improvements with baseline analysis and fail conditions.
CI/CD Ready
GitHub Actions, GitLab CI integration. JUnit XML output. Automated regression detection in your pipeline.
Python API
Use Heimr programmatically in your scripts. Import the Analyzer class and integrate analysis into any workflow.
Real Analysis Output
5-minute load test • 3,000 requests • Memory pressure scenario
# ❌ FAILED
**Reasons**: Anomalies: 108, Error/Warn Logs: 7
## Level 1: Primary KPIs
| Metric | Value | Reference |
|---|---|---|
| P95 Latency | 150.00 ms | < 500ms (API) |
| P99 Latency | 4456.00 ms | ⚠️ GC Pauses |
| Error Rate | 0.00% | < 1.0% |
| Throughput | 10.00 req/s | Stable |
## Level 2: Summary
- **Concurrency**: 5 VUs sustained for 5 minutes
- **Anomalies**: 108 detected (Avg 4021ms during GC pauses)
- **Prometheus**: CPU spikes 60-80% during GC events
- **Loki**: 7 GC pause warnings (1200-2500ms pauses)
- **Tempo**: Slow traces correlated with heap pressure
## Level 3: Per Endpoint Breakdown
| Endpoint | Requests | Error % | Avg (ms) | P99 (ms) |
|---|---|---|---|---|
| /api/checkout | 602 | 0.00% | 304.56 | 4129.39 |
| /api/inventory | 589 | 0.00% | 187.00 | 3618.20 |
| /api/orders | 617 | 0.00% | 228.38 | 4335.92 |
| /api/products | 588 | 0.00% | 275.53 | 4608.99 |
| /api/users | 604 | 0.00% | 285.47 | 4804.70 |