An open-source alternative to Scikit-Learn designed specifically for streaming data. It allows machine learning models to learn incrementally from one observation at a time without needing to store the entire dataset.
Tech Stack
PythonAI, ML & Data
Why river?
• No need for large training clusters
• Handles concept drift naturally
• Very low memory usage
Limitations
• Smaller algorithm selection than Scikit-Learn
• Not suitable for static batch data
• Single-threaded Python execution
3/6/2026
Last Update
607
Forks
114
Issues
BSD-3-Clause
License
Financial Leak Detected
Stop the "SaaS Tax"
Your team could be burning cash. Switching to river instantly boosts your runway.
Competitor Cost
-$1,440
/ year (est. based on Amazon SageMaker (Streaming))
