Stand Up a Data Platform for a Mobility-Data Startup's First ML Model
Overview
What this challenge is about.
As a 4-person team, build (1) a streaming ingestion path from a simulated telemetry source (Kafka + Python producer is fine); (2) a batch ETL job into a small warehouse (DuckDB or BigQuery sandbox); (3) a feature store slice (Feast or hand-rolled) serving 5 features; (4) a sample training job (any model is fine — the platform is the point) that pulls from the feature store; (5) tests + CI + infrastructure-as-code (Terraform or Pulumi). Produce a 6-page platform document plus a 30-minute handover for the incoming data engineer.
The Brief
What you'll do, and what you'll demonstrate.
Stand up a working data platform slice (streaming + batch + feature store + training job) for a mobility-data startup's first ML model.
Earning criteria — what you'll demonstrate
- Design a data platform across streaming, batch, and feature serving layers
- Implement infrastructure-as-code for reproducible environments
- Operate a feature store as the contract between data and ML
- Hand off a platform with documentation that survives without you
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
MLOps Engineer
Infrastructure-as-code plus feature-store contracts are bread-and-butter MLOps responsibilities.
This challenge sharpens
- infrastructure-as-code
- feature-store
- ci-cd
Machine Learning Engineer
Wiring a sample training job into the platform mirrors the MLE's daily integration work.
This challenge sharpens
- feature-store
- ci-cd
- team-collaboration