v0.5.8 · MIT License

The memory layer
your observability stack is missing

Siggy learns from every incident and fix. Three-layer retrieval makes recommendations sharper over time. Built on SigNoz.

> pip install siggy-memory copy
36ms avg response 23 failure types 20+ API endpoints 9 CLI commands

Why Siggy?

Modern observability gives you data. Siggy gives you answers.

Alert Fatigue

Teams drown in hundreds of alerts per day. Most are noise. The critical ones get buried because there is no historical context to prioritize them.

Tribal Knowledge

The fix for that Redis timeout lives in one engineer's head. When they are offline, the team restarts services blindly and hopes. Knowledge walks out the door every Friday.

Slow MTTR

Mean time to resolution has not improved in years because every investigation starts from scratch. No system remembers what worked before.

Everything you need to remember

Three-layer retrieval, explainable recommendations, and a feedback loop that gets smarter with every incident.

Memory Layer

Vector search via Qdrant, knowledge graph via SQLite, and Bayesian experience ranking. Three retrieval layers combined into one pipeline that improves over time.

Knowledge Graph

5 node types, 6 relationship types. Maps services, failure modes, and fixes into a connected graph that expands context during investigation.

Experience Engine

Records outcomes, computes success rates, and uses Bayesian ranking so the best fixes rise to the top. Every resolved incident makes future recommendations better.

Auto-Instrumentation

siggy instrument python app.py wraps any Python app with OpenTelemetry. Zero code changes. One command.

Explainable Recommendations

Every recommendation includes root cause, fix, confidence score, evidence chain, and reasoning. Not a black box -- a transparent assistant.

Zero-Config Setup

pip install siggy-memory then siggy quickstart. Embedded Qdrant (no Docker), local embedding fallback, no LLM required for normalization.

SigNoz Integration

Auto-creates 10-panel dashboards, 3 default alert rules, saved views, and webhook notification channels. Alert sidecar enriches in real time.

Alert Intelligence

Polls SigNoz alerts every 30 seconds, enriches with memory pipeline, and writes recommendations back as OTel spans visible in your SigNoz dashboard.

Demo Mode

siggy demo seeds 4 realistic incidents with outcomes. Works without SigNoz, without Docker, without anything but Python.

How Siggy works

Seven steps from alert to explainable recommendation.

Alert Fires

SigNoz detects an anomaly in your traces

Normalize

Incident becomes a structured KnowledgeObject

Vector Search

Qdrant finds similar past incidents

Graph Expand

Knowledge graph adds related services and patterns

Experience Rank

Bayesian scoring by historical outcomes

Explain

Root cause, confidence, evidence chain

Recommend

Actionable fix written back to SigNoz

Engineers record outcomes. The experience engine learns. Future recommendations improve.

Three-layer retrieval architecture

Every query flows through three memory layers. Each layer contributes signals. The result is ranked by real-world outcomes.

Incoming Alert Structured KnowledgeObject
1 Vector Search
Qdrant

Semantic similarity across all past incidents. Finds the closest matches by embedding distance.

2 Knowledge Graph
SQLite

Expands context through service dependencies, failure types, and fix relationships. 5 node types, 6 relationship types.

3 Experience Engine
Bayesian

Ranks candidates by historical success rate, resolution time, and engineer feedback. Best fixes rise to the top.

Explainable Recommendation Root cause · Fix · Confidence · Evidence chain

Deep SigNoz integration

Siggy does not just connect to SigNoz. It sets it up for you.

Auto-Created Dashboard

10 panels covering error rates, latency percentiles, memory hits, graph depth, experience scores, and recommendation throughput.

10 panels auto-created on startup
Default Alert Rules

3 alert rules configured out of the box. Catches high error rates, latency spikes, and high failure counts across all services.

3 default alert rules
Alert Intelligence Sidecar

Polls SigNoz alerts every 30 seconds, enriches through the memory pipeline, and writes recommendations back as OTel spans.

30s poll interval
Saved Views & Webhooks

Pre-configured saved views for incident investigation. Webhook notification channel for sending recommendations to Slack, PagerDuty, or any endpoint.

Auto configured on startup

Nine commands. One sidecar.

A CLI that works like a senior SRE sitting next to you.


        
siggy init Connect to SigNoz, configure memory
siggy quickstart One-command setup, auto-detects everything
siggy instrument Auto-instrument any Python app with OTel
siggy watch Alert sidecar, enriches with memory pipeline
siggy serve Start REST API server + sidecar
siggy up All-in-one: server + sidecar + your app
siggy investigate Manual investigation through memory pipeline
siggy status System health and connection checks
siggy demo Seeds 4 incidents, no SigNoz needed

Built for production

Battle-tested taxonomy, real benchmarks, measurable performance.

23 failure types
21 fix types
30 benchmark cases
20+ API endpoints
36ms avg response
5 graph node types
9 CLI commands
6 relationship types
10 dashboard panels
3 default alert rules
Python FastAPI Qdrant SigNoz OpenTelemetry SQLite Click Pydantic GROQ HuggingFace

Get started in 30 seconds

Three paths. Pick the one that fits.

Quickstart

Embedded Qdrant, auto-detects SigNoz and GROQ. Configures everything including dashboard and alert rules.

pip install siggy-memory && siggy quickstart
Step by step

Full control over each connection. Configure SigNoz URL, Qdrant endpoint, and LLM provider individually.

siggy init && siggy serve
Demo mode

Seeds 4 realistic incidents. No SigNoz, no Docker. Just Python 3.10+.

siggy demo

Requires Python 3.10+ and pip