retentioneering-tools
Python toolkit, MCP server, and agent skills for reproducible, auditable clickstream and event log analytics. Helps AI agents, data scientists and analysts buil
What is retentioneering-tools?
Python toolkit, MCP server, and agent skills for reproducible, auditable clickstream and event log analytics. Helps AI agents, data scientists and analysts build, validate, and cross-check product analytics, quantitative UX, customer journeys, graph-based user flows, behavioral segmentation, A/B tests, process mining models, Markov chain simulation
Overview of retentioneering-tools
What it is
Retentioneering is an open-source Python library for understanding user behavior from event logs. It helps find where users get stuck, compare paths different groups take, and discover patterns across clicks, sessions, and repeat visits. The library is designed for AI agents, data scientists, and analysts to build, validate, and cross-check product analytics, quantitative UX, customer journeys, graph-based user flows, behavioral segmentation, A/B tests, process mining models, and Markov chain simulation.
Key features
- Transition graph widget with interactive features: highlight routes, compare path groups, focus on steps, explore incoming/outgoing transition probabilities in ego view.
- Funnel widget for counting sessions that complete steps in order.
- Step matrix widget with diff mode for comparing behavior across segments.
- Step Sankey widget for visualizing flows around key events, with anchor specifications for precise alignment.
- Filter paths data processor using path metrics (e.g., length, duration, event count, pattern matching).
- Cluster analysis widget for interactive user clustering with silhouette scoring, heatmap, and segment saving.
- Segment creation and comparison across dynamic features (e.g., time periods, funnel stages).
Use cases
- Investigate metric changes by comparing journeys before/after a period.
- Identify where users leave a funnel and what they do instead.
- Analyze what leads to or follows an error or key event.
- Find sessions matching a specific behavior pattern.
- Explore behavioral clusters and compare product usage across user types.
Getting started
Python 3.10–3.13. Environment: Jupyter (Notebook, Lab, Desktop), VS Code, Cursor, Google Colab. Install the library: bash pip install retentioneering Within a notebook, use %pip install retentioneering. Quick test with bundled e-commerce dataset: python import retentioneering as rete stream = rete.datasets.load_ecom() stream.transition_graph()
License and pricing
Licensed under Apache-2.0. Open source, free to use.
Based on the project's public README and repository metadata.
