The Joe Reis Show

In today's Practical Data Modeling group discussion, we chatted about how to get buy-in for data modeling. The question was intentionally vague, because context is key. I give some thoughts on this topic, and how you can generalize this to most situations where you need to get buy-in. Practical Data Modeling: https://practicaldatamodeling.substack.com/

Vishnu Vasanth, with a deep technical vision, discusses analytical query engines, shifting left, the Indian tech scene, and the challenges and opportunities of AI impact on infrastructure. The conversation also touches on customer budget challenges, navigating the AI landscape, and the future of analytical query engines and object storage.

There's the interview you think you're going to have, then there's the interview you get. This is one of those, in the best way possible. I expected to chat about his time at Snowflake. We didn't even get past his early days building data warehouses because it was so fascinating. Did you know Kent is arguably one of the very first practitioners (probably an accidental inventor) of DataOps? This is sort of a "prequel" episode. Kent Graziano and I chat about his early days as a data practitioner.

Sometimes I feel like the data world is stuck in a world of tabular data (rows and columns). This has been the data world for decades. Let's think bigger. We've moved beyond data fitting into lakes. With the capability of AI to unlock the power of unstructured data (audio, images, video), it's time to start thinking about data oceans...

Keith Belanger, an OG data modeling practitioner, discusses the evolution of data modeling, agile methodologies, data mesh, and teaching conceptual data modeling. He shares insights on balancing agility with design, navigating data challenges in healthcare, and the excitement for data modeling innovations and AI impact.

This morning, the Practical Data Modeling Community held its first group discussion (to be posted very soon). People from all sorts of organizations (biggest companies in the world, universities, small companies) discussed how the approach analytical data modeling. My major takeaway - your mileage will vary. There's the ideal way of data modeling we're taught, and there's reality. Everyone's situation is different and there's no one-size-fits-all approach that will work for everyone. The discussion was awesome, and we'll do it again soon. If you're not part of the Practical Data Modeling Community, please join here: https://practicaldatamodeling.substack.com/

Kishore Aradhya and I both teach, and we agree this is a very difficult landscape to determine what and how to teach. Against the backdrop of generative AI, we discuss the role of universities in teaching tech and data, the role of a teacher, how to teach data, and much more. DSPY - https://github.com/stanfordnlp/dspy

Toby Mao, Founder of SQL Mesh, talks about simplifying data transformations. He discusses experimentation best practices, data tools, driving cars, and much more.

Matt Housley, a Data and ML engineer, discusses fads in data, data and ML engineering, tech hubs, and more. Topics include evolution of data tools and trends, balancing cutting-edge tech with foundational practices in ML, legal risks in business ventures, and navigating data ingestion costs for AI model training.

There's an inverse relationship between the value you add and how much you need to tell people about it. If you're adding value, you'll know - you don't need to talk about it. You're doing it. Also, the same goes with "data." If you're putting "data" as the center of the conversation, you just lost the game.