← Papers & projects

Technical report Systems research

Dynamic Prefill Optimization

Adaptive online packing for lower-TTFT LLM serving

Dynamic Prefill Optimization
−20%
Time to first token
p95
Online control signal
Greedy + DP
Packing strategies

This project introduces a workload-aware prefill optimization layer for LLM serving. It adapts batch timing from live tail-latency feedback and packs prompts by length before they enter a continuous batching engine.

Project Overview

  • Built an AIMD controller using real-time p95 time-to-first-token feedback.
  • Implemented greedy and dynamic-programming prompt packing to reduce padding waste across heterogeneous requests.
  • Integrated the controller as a lightweight queue layer before a vLLM-style continuous batching engine.
  • Evaluated on production-style DynamoLLM traces, achieving up to 20% lower TTFT.

Read the technical report or explore the implementation.