Design a Feed Generation for Capital One
Last updated: March 30, 2026
Quick Overview
Design a distributed feed generation system that handles millions of requests. Discuss trade-offs in consistency, availability, and performance.
Capital One
March 30, 202620
5
4,245 solved
Design a distributed feed generation system that handles millions of requests. Discuss trade-offs in consistency, availability, and performance.
This ML system design question from Capital One's Onsite tests your ability to think about ML systems at scale. The interviewer expects discussion of data quality, feature stores, model serving infrastructure, and A/B testing strategy.
What the Interviewer Expects
- Map the business problem to a concrete ML objective
- Propose reasonable features and a baseline model
- Discuss basic model evaluation metrics
- Outline a simple serving architecture
Key Topics to Cover
How to Approach This
- Start by clarifying functional and non-functional requirements with the interviewer.
- Estimate the scale: QPS, storage, bandwidth. This drives your design decisions.
- Draw a high-level architecture first, then deep dive into 1-2 critical components.
- Discuss trade-offs explicitly (e.g., consistency vs availability, SQL vs NoSQL).
- Address failure scenarios, monitoring, and how the system handles 10x traffic spikes.
Possible Follow-up Questions
- How would you run A/B tests on different model versions?
- What is your model retraining strategy?
- How would you ensure fairness and reduce bias in the model?
Practice a Similar Problem on Codemia
Solve a related problem with our interactive workspace, get AI feedback, and view detailed solutions.
Solve on CodemiaSample Answer
Requirements
Functional Requirements
- Real-time Feed Generation: The system should generate personalized financial insights and recommendations for millions of users in near real-time based on their tran...
Capacity Estimation
Assuming Capital One has around 50 million active users, if each user interacts with the feed generation system 10 times per day, this results in:
- Daily Requests: 50 million users * 10 interact...