We're sorry but this app doesn't work properly without JavaScript enabled. Please enable it to continue.

This lesson's interactive features are locked, please to keep using them

Large Language Models

Search has a problem: users aren't expert query-crafters.

  • "funny british bear" → Paddington
  • "that bear movie with the guy from titanic" → The Revenant

Large Language Models (LLMs) can help bridge this gap. LLMs are what ChatGPT, Claude, Gemini, and other new (but incredibly widely used) chatbots are based on. They're the hot new thing that's put AI back in the spotlight over the last few years.

LLMs are super useful in search – they can be used before or after the actual search takes place. In this chapter, we'll focus on using LLMs before the search to help improve the query itself.

  1. User enters a query.
  2. LLM processes the query:
    • Fixes typos
    • Expands meaning
    • Breaks apart complex queries
    • Adds missing context
  3. Search is performed with the improved query.
  4. Results are returned.