Overview
What this challenge is about.
You receive a pre-trained credit-risk model (a LightGBM model file) and a sample request payload. Containerize a FastAPI inference service, deploy to EKS or GKE (a single-zone cluster is fine), configure Horizontal Pod Autoscaler against CPU + a custom request-rate metric, and run a load test that ramps from 50 RPS to 500 RPS. Provide cost-per-1000-requests, p95 latency curves, and a 4-page runbook the on-call engineer will use.
The Brief
What you'll do, and what you'll demonstrate.
Move credit-risk inference onto autoscaling Kubernetes with sub-200ms p95 latency at 10x current load and a runbook the on-call team can use.
Earning criteria — what you'll demonstrate
- Containerize and deploy a model service to a managed Kubernetes cluster
- Configure horizontal pod autoscaling against a custom metric
- Conduct a realistic load test and interpret latency curves
- Write a runbook an on-call engineer can use under pressure
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
Kubernetes-native model serving with autoscaling and an on-call runbook is the day-one work of an MLOps engineer at any fintech or large-AI company.
This challenge sharpens
- kubernetes
- model-serving
- monitoring-design
AI Engineer
Closing the loop from container to production inference under SLA is the AI-engineer skillset that ships product features.
This challenge sharpens
- containerization
- model-serving
- load-testing
AI Solutions Architect
Designing autoscaling inference platforms with cost + latency trade-offs documented is the core craft of an AI solutions architect.
This challenge sharpens
- kubernetes
- autoscaling
- monitoring-design