Building a Search Chatbot with GPT-3 (Part 0)
One of the biggest problems workers at big companies face is finding the information needed to do their job. If you’ve worked at any mid or large size company, then you’ve probably run into the following problem at some point or another:
- You need to find some piece of proprietary company information
- You search the company wiki
- You find the information… an hour later
Knowledge about internal systems and processes is often buried among hundreds or thousands of unrelated documents. Keyword searches often give poor results, so it can be next to impossible to find the piece of information you need. Luckily, the power of AI can dramatically improve our ability to find this information quickly and efficiently.
In this series, I’ll provide a simple, high level overview of how to use OpenAI’s GPT-3 tool to improve knowledge retrieval. Though I won’t dive into code-level details, I’ll provide some helpful tools and considerations to think through when building these systems.
- Part 0 — Overview
- Part 1— How Does GPT-3 Work with Search?
- Part 2— Implementing Semantic Search
- Part 3 — Integrating with GPT-3
If this sounds like something you’re interested in, please stay tuned for the rest of the series!