Browse categories
Explore
Fiverr Pro
English
$
USD
I will coach you for an ai or ml system design interview at a top lab
Senior AI Researcher and Software Engineer
I am a Senior AI engineer working on production LLM systems, currently building generative AI applications for FWD Insurance. I author a newsletter read by 16,000+ engineers covering LLM system design...
About this Gig
I wrote 1,141 pages on RAG interviews and a full guide to LLM system design interviews, and I publish to 20,000+ engineers preparing for exactly these loops. I have spent a year mapping where senior AI candidates lose the offer.
It is almost never the architecture. It is the follow-up question, the trade-off you skipped, and the number you could not produce out loud.
What you get:
- A mock round on the topic you choose: LLM system design, RAG, inference, agents, or classic ML system design
- Pressure on your answer the way a senior interviewer applies it, not a friendly walkthrough
- Written feedback naming the specific patterns that read as junior
- The arithmetic you are expected to do in the room: KV cache sizing, index memory, scaling laws, cost per token
- What to say when you do not know, which is a real interview skill most candidates get wrong
I will tell you where you actually are. If you are not ready for the loop you are aiming at, you will hear that, with the gap named and an order to close it in.
Message me with the role and the company and I will tell you what to expect.
Interview type:
Job
Experience level:
Associate
•
Senior
Industry:
Internet & Technology
•
Engineering
Company expertise:
•
Databricks
•
Nvidia
•
OpenAI
•
Scale AI
Session focus:
Mock interview
Language:
English
•
Vietnamese
FAQ
Is this a real mock or a friendly walkthrough?
A real mock. I interrupt, I push on the trade-off you skipped, and I ask for the number. Friendly practice is comfortable and it does not find the thing that is costing you offers.
Which topics can you cover?
LLM system design, RAG and retrieval, inference and serving, agent systems, and classic ML system design. Pick one topic per round so it goes deep instead of wide.
What if I am not ready yet?
You will hear that, with the gap named and an order to close it in. I would rather tell you before the real loop than agree with you after it.
Can you tell me what my target company asks?
I can tell you what that class of loop tests and how the questions read, from public material and from what my readers report back. I do not have leaked questions and I would not use them.
