ML System Design: CV interview questions

Design a face recognition system for real-time use with cameras in public spaces. The system should handle 5000 concurrent users and process 1000 images per second, with a response time of no more than 2 seconds.

You need to build a product image search system based on text queries. The system should support 1 million users and handle 5000 queries per second with a latency of no more than 1 second.

Design a video surveillance monitoring system for large public events. The system should process 2000 video streams simultaneously with minimal latency and ensure data storage for 30 days with a data volume of about 10 TB per day.

Build a medical imaging analysis system that can process 100,000 images per day. The system should provide an accuracy of at least 95% and a response time of no more than 5 seconds per image.

You need to design an OCR system supporting 10 languages. The system should process 50,000 documents per day with an accuracy of at least 90% and a response time of no more than 3 seconds per document.

Design a perception system for autonomous driving. The system should process information from 10 sensors at a frequency of 20 Hz and ensure a response time of no more than 100 milliseconds for critical decisions.

You need to build a quality inspection system for a production line. The system should process 2000 images per hour and ensure defect detection with an accuracy of at least 95%. The budget for implementation is $100,000.

Build a face detection system in a video stream that must process 100 video streams in real-time with a maximum latency of 200 milliseconds. The system should operate on hardware with a budget of up to $50,000.

Design a product search system based on images that supports 500,000 users. It should handle 10,000 queries per second and return results with a latency of no more than 1 second.

Build a medical analysis system using MRI images. The system should process 1,000 images per day and provide an accuracy of at least 98%, with a response time not exceeding 4 seconds.

Design a face recognition system that needs to handle 100,000 users simultaneously. The system should provide a response time of less than 2 seconds and be capable of processing 500 queries per second. How would you design this system?

You need to build a product image search system that can handle 10,000 queries a day. The system should support 3,000 unique images and provide search results with a latency of less than 1 second. How would you approach this task?

Design an OCR system that needs to process 200 documents per minute with a maximum accuracy of 95%. The system should be able to scale up to 10,000 documents per minute within the next five years. How would you design this architecture?

You need to design a video surveillance system that should process and analyze 4K video streams with a latency of less than 5 seconds. The system should support up to 50 cameras simultaneously. What technologies and architectural solutions would you use?

Design a medical imaging system that needs to process 1 TB of images per day. The system should allow access to images for 100 doctors simultaneously with a latency of less than 2 seconds. How will you address data storage and processing challenges?

You need to build a perception system for autonomous driving that must process data from 10 sensors in real-time, ensuring a latency of less than 20 ms. The system should be able to handle up to 1,000,000 data records per second. What are your key considerations in designing this system?

Design a quality inspection system that needs to process 5,000 products per hour and detect defects with an accuracy of at least 90%. The system should provide a response time of less than 3 seconds for each inspection. What approaches would you use to achieve this?

You need to design a face recognition system that operates under low light conditions, processing 1,000,000 images with an expected accuracy of 85%. The system should provide processing time of less than 1 second for each image. How will you handle these constraints?

Design a text-based image search system that should handle 50,000 queries a day and provide results in less than 2 seconds. The system should support 10 million images. What are your key architectural decisions?

You need to build a system for real-time analysis of 100 video streams, ensuring a latency of less than 10 seconds for each stream. The system should support scaling up to 5,000 streams in the future. How will you ensure system performance and reliability?