“How AI Works in Business Intelligence Tools" Opening the black-box by Aria Voron."The presentation breaks down shows what happens behind the scenes when someone types a question like “What were Q1 sales for Scranton?” into a tool like Tableau Pulse (or any other similar BI AI chat bots), AI starts a multi-step process behind the scenes. First, it cleans and corrects the text, then breaks it into tokens to identify key elements like dates, metrics, and locations. Using models trained on analytical language, it detects intent (e.g. “show sales”) and maps those terms to actual fields in the company’s data model — like turning “sales” into sales_amount and “Q1” into a specific date range. From there, it auto-generates a SQL query, runs it against the data warehouse, and formats the results into a readable chart, KPI card, or summary sentence. The system learns from user feedback over time, refining its understanding of business-specific terminology and logic.What used to take an analyst hours now takes seconds — and the answer appears in plain English.This is mostly the content of the presentation. It also shows how by using the AI you can learn something that would require you to go back to school and get another degree in computer science - now with the help of ChatGPT or Gemini could be explained in a few hours..This concept, the framework of leveraging AI could be extrapolated on so many other topics.