Overview
What this challenge is about.
Pull AWS Cost and Usage Report, GCP billing export, and Lambda Labs invoices into a single Parquet table. Implement a tagging convention (project + client + experiment_id) and a 'cost attribution' step that handles untagged spend. Build a dashboard (Streamlit, Metabase, or Looker Studio) with: per-client weekly spend, top-10 most expensive experiments, and an untagged-spend alert. Deliver the dashboard, the underlying pipeline, and a 3-page partner-team rollout memo.
The Brief
What you'll do, and what you'll demonstrate.
Ship a per-client GPU cost dashboard that flags unattributed spend and lets partners catch unprofitable engagements within a week.
Earning criteria — what you'll demonstrate
- Reconcile billing data across multiple cloud providers
- Design a cost-attribution model that handles tagging gaps
- Build a partner-facing dashboard with action-oriented views
- Document a tagging convention engineering will actually adopt
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.
Data Engineer
Multi-cloud billing reconciliation + dashboard is the kind of pragmatic data-engineering project that ships value in week one of a new job.
This challenge sharpens
- etl-pipelines
- data-modeling
- cloud-cost-attribution
MLOps Engineer
Cost attribution for GPU experiments is MLOps-adjacent work that keeps research budgets honest at any AI company.
This challenge sharpens
- cloud-cost-attribution
- etl-pipelines
- dashboarding
AI Product Manager
Designing partner-facing cost views with clear action triggers is the AI PM's craft of turning data into decisions.
This challenge sharpens
- dashboarding
- documentation
- data-modeling