AI System Design Practice

A 60-minute hands-on practice session focused on designing AI and ML systems: the kind of problems you'll face in interviews and on the job.

How it works:

1. We pick a real-world AI system design problem together
2. You walk through your approach while I ask follow-up questions and challenge your assumptions
3. I share how these problems play out in production based on my experience shipping AI at Samsung (Galaxy AI, on-device LLMs) and 1mind (enterprise knowledge systems)
4. We wrap up with concrete areas to strengthen

Topics we can cover:

1. Knowledge retrieval and RAG system design
2. NLP pipeline architecture (ASR, translation, search ranking)
3. On-device vs. cloud ML trade-offs
4. AI product design end-to-end (from requirements to deployment to evaluation)
5. Real-time inference and serving architectures
6. Evaluation frameworks for LLM-based systems

Why practice with me:

1. I've built these systems in production: on-device LLMs for Galaxy phones, knowledge pipelines for enterprise AI, evaluation frameworks that replaced manual QA.
2. I can tell you what works on a whiteboard vs. what actually survives production.

This is collaborative practice, not a scored exam.

Best for: Engineers preparing for ML system design interviews or looking to think more architecturally about AI systems.