I've been on both sides of this. Interviewed at Deloitte a while back and have since talked to a bunch of people who went through it for senior and staff-equivalent roles in their tech practices. Here's the honest picture on system design.
First: Deloitte is not Google. The system design round (when it exists) is calibrated to consulting delivery context, not pure product-company scale problems. This surprises people coming from FAANG prep.
When system design actually shows up Not every loop has it. It's more common for: senior consultant (SWE equiv), manager-level tech roles, and specialist roles in cloud architecture or data engineering. Pure implementation roles may skip it entirely.
What they actually ask Expect pragmatic design questions. Not "design Twitter." More like: Design a document processing pipeline for a government agency How would you architect a reporting layer on top of a client's existing ERP Walk me through how you'd build a secure data ingestion pipeline for a healthcare client
They care about: security posture (data classification, access control), cloud provider knowledge (AWS, Azure, GCP), integration with legacy systems, and whether you think about the client's actual constraints.
What they don't care about (vs FAANG) Horizontal scaling to millions of QPS is not the core question. They're less interested in you knowing exactly what Kafka does at 10M msg/s and more interested in whether you can explain why you'd choose Kafka vs SQS vs a simpler queue given a client's existing stack.
How to prep If you're coming from a product company, dial back the scale obsession. Lead with requirements gathering, security considerations, and trade-offs given real-world constraints. Practice explaining architecture to a non-engineer. That actually comes up.
The interviewer is often a project manager or delivery lead sitting in alongside the tech lead. Your explanation needs to land for both audiences.