Build a production ML system with only Python on free serverless services
summary
An end-to-end Python system predicts PM2.5 for Seattle by scraping live sensors and weather forecasts, writing features to Hopsworks, training an XGBoost model on pandas-derived feature views, and running daily batch inference on Modal with a Streamlit UI deployed to Hugging Face Spaces. The pipelines include point‑in‑time joins, Great Expectations checks, model registry saves, and automated scheduling; dozens (≈40–50) similar serverless stacks are already running and the demo system keeps collecting data and producing usable hindcasts.
AI-written summary · about
chapters (8)
- 0:00Serverless ML systems
- 10:30Data scientists and prediction services
- 21:42Free serverless ML tools
- 36:15Feature pipeline with Python
- 56:17Model training and artifacts
- 1:01:12Inference pipeline
- 1:10:00MLOps for production
- 1:17:26Student projects and ideas
related talks
41mBuilding Serverless Machine Learning Models in the Cloud
31mHow to build a serverless electricity price prediction service in Python with Hopsworks + Streamlit
36mBuild prod ML systems with feature /training /inference pipelines
44mML in Production
37mAutomated Machine Learning: Building a Pipeline
45mWorkshop: Deploy and Monitor ML Pipelines with Open Source and Free Applications
17mR & Python playing nice, in Production
25mML model from an idea to production with the help...