Role track

ML Engineer

Train, serve, and rollback. Features as tables. Latency as an SLA.

155 questions · Python · Features · Serving · System design

Suggested sequence

1. Python / tensors
The same problem in NumPy and PyTorch.
MLE coding →
2. The flowchart
Features, serving, and how you knew it got worse.
MLE roadmap →
3. AI system design
Batch vs online. Warehouse is the wrong online store.
System design →
4. Eval and incidents
How you knew the model got worse.
AI / ML questions →

medium · Warehouse · Capital One
A Medium warehouse is slower but cheaper than Large. When is that true?
Practice →
medium · Warehouse · Snowflake
Someone DELETED gold.orders. What do you do in the first ten minutes?
Practice →
medium · Warehouse · McKinsey
Result cache vs warehouse cache vs metadata cache — which one did this dashboard hit?
Practice →
medium · Warehouse · Palantir
Would you put Cortex COMPLETE inside a Looker explore?
Practice →
medium · Warehouse · Netflix
Micro-partitions and clustering: when does a clustering key pay for itself?
Practice →
medium · Warehouse · Airbnb
Streams and Tasks vs an external orchestrator. How do you choose?
Practice →