← Back to Directory

Credit Karma

Large Enterprises

Screna has 4 Credit Karma interview write-ups. The most reported round is the system design round (3).

4 interview notes · updated Jul 2026

Credit Karma·Machine Learning Engineer·Onsite - System Design / Architecture

Jun 2026
Went through a system design round at Credit Karma focused entirely on large-scale ML model serving. Seven questions and they were all variations on the same core problem: how do you serve thousands of models to hundreds of millions of users without everything falling apart. Tough round, lots of follow-ups.
  • How would you design a serving architecture capable of handling hundreds of millions of users across thousands of models?
  • At serving time, how do you route an incoming request to the correct model out of thousands of options?
  • What strategies would you use to efficiently manage and serve thousands of models simultaneously?
  • What are the tradeoffs between keeping models resident in memory versus loading them from disk on demand?
  • How would you design a scalable model-loading strategy for a system with thousands of models?
  • How do you differentiate handling between frequently used (hot) models and rarely used (cold) models in a serving system?
  • How do you reason about the tradeoff between memory utilization and serving latency in a large-scale ML system?

“This was the opening question and it set the tone.”

View Post

Credit Karma·Machine Learning Engineer·Onsite - System Design / Architecture

May 2026
Onsite system design loop at Credit Karma for an ML Engineer role. The whole session revolved around one meaty multi-task ranking problem with a bunch of follow-ups layered on top. Technically pretty intense, lots of debugging-under-pressure vibes.
  • You're training a multi-task ranking model that predicts click, application, conversion, and approval. Training loss goes NaN after a few hundred steps. Walk through how you'd debug it.
  • Offline metrics look great but the model underperforms in the live experiment. How do you investigate the offline-online gap?
  • Conversion and approval labels have a positive rate around 0.1%. What loss functions and optimization strategies make sense for such extreme imbalance, and how does label delay factor in separately?
  • A high-importance feature is always a default constant at serving time but is correctly populated in the offline training data. What do you do, and how would you have caught this earlier?
  • Mixed precision training is causing instability but you need it for throughput. How do you keep both?
  • Conversion labels take 14 days to mature. How does that change your loss construction and your offline evaluation setup?
  • How would you set up continuous monitoring to automatically catch training-serving skew before it ships?

“The thing that tripped me up initially was wanting to list every possible cause (gradient explosion, bad LR, loss weights, etc.) before actually thinking about what the timing tells you.”

View Post

Credit Karma·Machine Learning Engineer·Technical Phone Screen

May 2026
Interviewed for an ML Engineer role at Credit Karma and the technical portion leaned heavily into model training diagnostics. The questions were meaty and definitely not the kind you can wing with surface-level prep.
  • What are the possible reasons a multi-task model's loss goes NaN or becomes unstable after initially decreasing?
  • How would you debug training instability that only shows up after a few hundred steps rather than at the start?
  • What loss functions and optimization strategies would you use for a sparse conversion prediction problem?

“This one tripped me up more than I expected.”

View Post

Credit Karma·Machine Learning Engineer·Onsite - System Design / Architecture

Apr 2026
ML system design loop for Credit Karma, focused entirely on recommendation and ranking for their financial product marketplace. Heavy on the ML infra side, lots of follow-ups about label sparsity and multi-stage architectures. Felt more like a research discussion than a standard interview.
  • How would you design an ML system for Credit Karma to maximize revenue from financial product recommendations?
  • Should you use a single model across the entire funnel or separate models for each stage like CTR, application rate, and approval rate?
  • How do you handle conversion labels that arrive days or weeks after the original impression?
  • Conversion rates in this domain are extremely sparse, around 0.1 to 0.5 percent. How do you deal with that during training?
  • If you apply negative sampling during training, how do you correct for that bias at inference time?
  • You're training on multiple objectives like CTR, conversion, and approval rate. How do you combine those into a final ranking score?
  • What ranking objective function would you actually use to combine multi-objective scores into a final ordering?
  • Compare a two-stage retrieval plus reranking architecture to scoring all candidate items directly in a single pass.
  • What are the tradeoffs between a two-stage and a single-stage ranking architecture?
  • Under what conditions would you choose a two-stage architecture over a single-stage one?

“Big open-ended opener.”

View Post