LLM-Native Recommendation Systems

Understand how LLMs are replacing traditional recommendation ML infrastructure — from feature stores to prompt design, two-phase training to prefill-only serving — and what that means for building modern personalization systems

3 sections · 8 lessons

Course outline

From Feature Engineering to Context Engineering

  1. How traditional recommendation systems use feature engineering
  2. What context engineering means in LLM-native systems
  3. Designing compact, high-information prompts

Training LLM-Native Rankers

  1. Two-phase training: foundation adaptation vs. task fine-tuning
  2. Balancing ranking, language modeling, and alignment objectives
  3. Data efficiency advantages of foundation models

Serving LLMs at Recommendation Scale

  1. Prefill-only inference and why it matters for cost
  2. Infrastructure shift: feature stores to LLM serving stacks

Start learning with Wondering