Overview
Explore a novel time-series processing pipeline for MLOps in this 22-minute talk from the Data Science Festival Oktoberfest 2023. Discover how Piotr Stepinski from Infinitii ai combines Apache Arrow with S3 for handling large-scale file processing, utilizes Polars or DuckDB for data manipulation, and implements a high-speed task runner written in Rust. Learn about WASM tasks for forecasting, with a focus on the current state-of-the-art Python in WASM implementations. Gain insights into potential performance improvements for production deployments of time-series models. Suitable for those with some technical knowledge, this introductory-level presentation offers valuable insights into cutting-edge data science techniques and tools.
Syllabus
Forecasting time series with Arrow and WASM
Taught by
Data Science Festival