Finished the Snowflake SWE loop last week for a senior IC role on their query optimization team. Sharing the full breakdown because I couldn't find anything current when I was prepping.
Total timeline: applied via LinkedIn, heard back in 8 days, offer 5 weeks after first contact.
Round 1: recruiter screen (30 min) Standard stuff. Current role, why Snowflake, availability. The recruiter was sharp and actually knew the team's technical scope. She mentioned Snowflake prioritizes candidates who've worked at scale, so dust off your biggest numbers.
Round 2: technical phone screen (60 min) One coding problem, medium difficulty, graph-adjacent. I used Python. The interviewer ran it in a shared IDE (Coderpad). They also asked one follow-up about time complexity and what happens at 10x scale. Nothing brutal but no free rides either.
Round 3: online assessment (separate from phone screen in my case) Two LeetCode-style problems, 90 minutes. One medium, one that felt medium-hard. Timed. I'd say 70% of the difficulty is just not panicking.
Onsite (virtual): 5 rounds over two days 2x coding (one graph, one on strings/parsing) 1x system design (distributed query execution, very on-brand for Snowflake) 1x behavioral 1x hiring manager chat
The system design was the most Snowflake-specific thing. They clearly want you to understand multi-tenant architecture, caching layers, and what makes columnar storage fast. If you haven't read up on how Snowflake's virtual warehouses actually work, do that before you walk in.
Total: took about 3.5 hours across both days. Interviewers were friendly but thorough. Debrief took 6 business days, which felt long but the recruiter kept me updated.
Happy to answer questions. Offer came in at the lower end of what I expected but the recruiter said there was room to move.