This workshop aims to teach basic Feast concepts and walk you through focuses on how to achieve a common architecture:

TODO: add architecture diagram

• Data sources: Kafka + File source
• Online store: Redis
• Use case: Predicting churn for drivers
  • Batch scoring via offline store
  • Real time scoring via online store

We will generate a model that will predict whether a driver will churn.

This workshop assumes you have the following installed:

• A local development environment that supports running Jupyter notebooks (e.g. VSCode with Jupyter plugin)
• Python 3.7+
• pip
• Docker & Docker Compose (e.g. brew install docker docker-compose)

First, we install Feast with Redis support:

pip install "feast[redis]"

We then use Docker Compose to spin up a local Kafka cluster and automatically publish events to it.

• This leverages a script (in kafka_demo/) that creates a topic, reads from feature_repo/data/driver_stats.parquet, generates newer timestamps, and emits them to the topic.

docker-compose up

Install Feast using pip

pip install 'feast[redis]'

We have already set up a feature repository in feature_repo/.

Deploy the feature store by running apply from within the feature_repo/ folder

cd feature_repo/
feast apply

Output:

Created entity driver
Created feature view driver_hourly_stats
Created feature view driver_daily_features
Created on demand feature view transformed_conv_rate
Created feature service convrate_plus100

Deploying infrastructure for driver_hourly_stats
Deploying infrastructure for driver_daily_features

Next we load features into the online store using the materialize-incremental command. This command will load the latest feature values from a data source into the online store.

CURRENT_TIME=$(date -u +"%Y-%m-%dT%H:%M:%S")
feast materialize-incremental $CURRENT_TIME

Output:

Materializing 2 feature views to 2022-04-28 12:38:01-04:00 into the redis online store.

driver_hourly_stats from 1748-07-13 16:38:03-04:56:02 to 2022-04-28 12:38:01-04:00:
100%|████████████████████████████████████████████████████████████████| 5/5 [00:00<00:00, 759.42it/s]
driver_daily_features from 1748-07-13 16:38:03-04:56:02 to 2022-04-28 12:38:01-04:00:
100%|███████████████████████████████████████████████████████████████| 5/5 [00:00<00:00, 1081.28it/s]

Now run the Jupyter notebook (feature_repo/workshop.ipynb)