• Timeslot: Sunday 6th April 2025, 16:00-17:00, Room C
• Tags: AI

How does struggling with AI product development in the Python landscape look today? Any IT professional knows the perils of using recursion, but we’ve taken a bold step forward—leveraging an AI-driven tool, created by us, to create AI-powered products. Despite the challenges, we’ve emerged stronger. Join me for a talk where we’ll demystify data classification using language models (LLMs) and the Snowflake platform, all powered by Python. Learn how our team’s innovative data product has transformed data governance and security. * Why did we decide to build our own Data Classification product? * How do we define success for the project? * What led us to select the right LLM model that ensures confidentiality and propels our progress? * How do we deal with setbacks, pivoting the product, and pushing ahead amid the uncertainties around LLM models? * What’s next in securing our data?

Experience a session where we showcase the magic of automatic data classification for sensitive information, such as PII and MNPI, turning data chaos into clear insights.

I promise - no fluff, no buzzwords. Solely a first-face story from the GitLab Data Team member, created from sweet and pain from the world leader of DevSecOps platform creator and AI-based tool, remote work champion and one of the most transparent company in history.