Don’t start building your own LLMs before you master the basics.
Everybody is talking about AI and how it will change business.
Most people are thinking about creating some kind of ChatGPT-like Large Language Model application.
But when asked for a specific use case (or even a business model), few actually have one.
Here are five specific things you can focus on in the meantime:
-
LLM Use Cases: Do you have one - or even better, several - strong use cases for an LLM? Or are you just experimenting to keep up with the cool kids?
-
Strategy: Do you have a data strategy in place? And how do AI and LLMs fit into it?
-
Data Platform: Is your data ready to use on a modern data platform like Databricks, Snowflake, or another lakehouse or data warehouse? Without data context, LLMs are practically useless.
-
Data Products: Have you already implemented use cases that deliver business value? Think of foundational data products like meaningful reports or ad-hoc analytics that support business decisions.
-
Strong Team: Do you have the right people with the skills to build data products that drive your business?
Invest in the basics and start creating value from your data. LLM use cases will emerge naturally along the way.
