Mei Lin

Author

@meilin

ML Infrastructure Engineer

Serving large models efficiently. KServe, vLLM, and the dark art of batching.

Singapore meilinJoined Jul 7, 2026

Stories

182

Views

Likes

Followers

JanFebMarAprMayJunJul

11 contributions in the last 6 monthsLessMore

How vLLM's paged attention and dynamic batching transformed our inference economics.

Jul 51 min4610ML Inference

Draft models, acceptance rates, and the latency wins that survived contact with reality.

Jun 31 min1366ML Inference