How to build a serverless electricity price prediction service in Python with Hopsworks + Streamlit
summary
Builds a serverless electricity price prediction service in Python and actually wires it end to end: Pandas feature engineering feeds a Hopsworks feature store using a Python DSL that guarantees point‑in‑time joins and skew‑free transforms, Great Expectations enforces feature quality and rejects bad writes, training produces a saved model and hourly precomputed predictions land in the online store, and a Streamlit dashboard reads online features to display results so the pipeline demonstrably runs and serves predictions.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 2:10Why predict Swedish prices
- 5:40Serverless architecture breakdown
- 7:26State management with Hopsworks
- 10:39Feature store fundamentals
- 14:11Joining features correctly
- 16:40Transformations and normalization
- 18:48Data validation with Great Expectations
- 22:19Model training and predictions
- 24:49Streamlit UI for predictions
- 27:19Q&A
related talks
01h 23mBuild a production ML system with only Python on free serverless services
23mHopsworks Feature Store after 4 years: Lessons learned and what's next
44mShaping the Future of Real-time Data Pipeline
38mBuild a personalized Commute virtual assistant in Python w/ Hopsworks & LLM Function Calling
01h 02mUsing Python to Power Serverless Applications
45mWorkshop: Deploy and Monitor ML Pipelines with Open Source and Free Applications
46mServerless for Data Scientists
15mKick-off