Ho Bae joined Ravit Jain on The Ravit Show to talk about what makes enterprise data AI-ready.
Most companies already have plenty of data. What they lack is data an AI task can actually use. Ho also walked through a live demo, tracing two network alerts back to the same device without the model ever seeing the original device name.
What they talked about
- 01Why having data and having AI-ready data are two different things
- 02Why AI-ready depends on the task, not data in general
- 03Why removing too much data can break the AI by stripping out relationships
- 04How LLM Capsule returns results inside the workflow with the real business values
- 05Keeping your current model, and starting with one blocked AI task
