LLM Inference Mock Interview + Feedback with Harshal Dharpure

A realistic technical round for ML engineer, LLM engineer and inference or performance roles: transformer and attention basics, KV cache and memory math, quantization, batching, speculative decoding and latency vs throughput trade-offs, plus one short design or reasoning problem. The last 15 minutes are honest feedback, and within 24 hours you get a written scorecard with the 3 things to fix first.