Feed an AI-agentic beast with proper data

In this hands-on workshop we’ll work through four real-world scenarios in Python: finding the minimum viable context for sales questions, keeping frequency signals intact when summarising support tickets at scale, selecting the right columns for churn analysis on a wide customer table, and extracting structured fields from messy product listings with validation you can trust. You’ll write and discuss code together — practical patterns you can take straight back to work. Timeline: 00:00:20 - PyLadies Amsterdam introduction 00:01:55 - Jessica Eggen introduction 00:05:08 - Workshop topic introduction 00:07:23 - Workshop setup and exercise 1 - Scrub before you send: removing PII from data + time for exercise 1 00:32:31 - Exercise 1 Walk-through 00:34:40 - Exercise 2 - Minimum viable context + time for exercise 2 00:58:59 - Exercise 2 Walk-through 01:00:28 - Exercise 3 Right context, right question + time for exercise 3 01:16:12 - Exercise 3 Walk-through 01:18:05 - Exercise 4 - When the LLM transforms the data + time for exercise 4 01:25:45 - Exercise 4 Walk-through 01:27:14 - Exercise 5 - Numbers they can trust 01:27:31 - Workshop wrap-up 01:28:31 - PyLadies Amsterdam and community announcements GitHub Repo https://github.com/pyladiesams/feed-a... Speaker:   / jessica-eggen-579b50155   Jessica Eggen is a Data Analytics Engineer who loves to spent her time wrangling and puzzling with real-world data at scale. With a background in Business Analytics and machine learning she hopes to bring a practical, business-grounded perspective to AI.