ChatGPT, Gemini, Copilot – LLMs are everywhere. But how do you integrate them into your operations? After a year spent developing our own LLM, the Zero100 team shares hard-won lessons and a primer on getting LLM design right. Featuring: VP Data & Analytics Cody Stack, Senior Manager, Data Science Julia Dahlgren, and Data Scientist Prajwal Bn.
- Why we built our own LLM (1:37)
- Key considerations for your build vs buy decision (2:55)
- How we leveraged open-source components (5:39)
- The promise – and challenge – of knowledge graphs (7:42)
- Encouraging adoption and winning buy-in from leadership (9:38)
- Why prompt engineering and quality training data are underrated (12:44)
- The business case for building a small language model (14:46)