streamframe is a lightweight engine for computing real-time features over streaming data.
It’s built in Rust for performance, but designed for ease of use via Python.
Most tools like pandas or Polars assume you already have all your data.
But in many real-world systems, data arrives continuously:
- API requests
- logs
- sensor readings
- market data
You don’t want to recompute everything every time a new value comes in.
streamframe updates everything incrementally, in constant time.
pip install streamframe