Stay ConnectedI’m Joe Reis, co-author of Fundamentals of Data Engineering, creator of Mixed Model Arts, international speaker, and long-time observer of the tech industry’s favorite recurring mistakes.
Weekly broadcasts from somewhere on the road. Expect raw solo rants and long-form conversations with the engineers and leaders actually building things. I dive into data, AI, architecture, and career survival.
This podcast is invitation-only. Unsolicited guest pitches and PR outreach are not considered.
Copyright: © Joe Reis
In this episode, I sit down with Ole to discuss his new book, "Fundamentals of Metadata Management." We move past the simple definition of "data about data" to a more nuanced view of metadata as something that exists in two places at once , serving as a pointer to find information elsewhere.
Ole introduces his core concept of the "MetaGrid"—the interconnected, yet siloed, web of metadata repositories that already exists within every large organization across various teams and technologies. He argues that the key to better metadata management is not to build a new monolithic system but to recognize, document, and integrate the MetaGrid that's already there, hiding in plain sight.
The conversation also covers the impact of the AI hype cycle , the lessons learned from the Data Mesh movement , the sociological incentives that help or hinder metadata projects , and the cultural clash between the worlds of data engineering and library science.
The way we work is changing right in front of us. In this rant, I talk about how I'm seeing AI reshape how technical and non-technical people do their work. The bottom line - there's a lot of room to innovate and evolve your job.
In this discussion, I sit down with data veterans Remco Broekmans and Marco Wobben to explore why so many data projects fail. They argue that the problem isn't the technology, but a fundamental misunderstanding of communication, culture, and long-term strategy.The conversation goes deep into the critical shift from being a "hardcore techie" to focusing on translating business needs into data models. They use the classic "involved party" data modeling pattern as a prime example of how abstract IT jargon creates a massive disconnect with the business.Marco shares a fascinating (and surprising) case study of the Dutch Railroad organization, which has been engaged in an 18-year information modeling "program" - not a project - to manage its immense complexity. This sparks a deep dive into the cultural and work-ethic differences between the US and Europe, contrasting the American short-term, ROI-driven "project" mindset with the European capacity for long-term, foundational "programs".Finally, they tackle the role of AI. Is it a silver bullet or just the latest shiny object? They conclude that AI's best use is as an "intern" or "assistant", a tool to brainstorm, ask questions, and handle initial prototyping, but never as a replacement for the deep, human-centric work of understanding a business.Timestamps:00:00 - Introduction01:09 - Marco Wobben introduces his 25-year journey in information modeling.01:56 - Remco Broekmans reintroduces himself and his focus on the communication aspect of data.03:22 - The progression from hardcore techie to focusing on communication over technology.08:16 - Why is communication in data and IT projects so difficult? 09:49 - The "Involved Party" Problem: A perfect example of where IT communication goes wrong with the business.13:35 - The essence of IT is automating the communication that happens on the business side.18:39 - Discussing a client with 20,000 distinct business terms in their information model.21:55 - The story of the Dutch Railroad's 18-year information modeling program that reduced incident response from 4 hours to 2 seconds.27:25 - Project vs. Program: A key mindset difference between the US and Europe.34:18 - The danger of chasing shiny new tools like AI without getting the fundamentals right first.39:55 - Where does AI fit into the world of data modeling? 43:34 - Why you can't trust AI to be the expert, especially with specialized business jargon.47:18 - The role of risk in trusting AI, using a self-driving car analogy.53:27 - Cultural differences in work pressure and ethics between the US and the Netherlands.59:29 - Why personality and communication skills are more important than a PhD for data modelers.01:03:38 - What is the purpose of an AI-run company with no human benefit? 01:11:21 - Using AI as an instructive tool to improve your own skills, not just to get an answer.01:14:12 - How AI can be used as a "sidekick" to ask dumb questions and help you think.01:18:00 - Where to find Marco and Remco online
Are you a giver or a taker?
It seems like every few months, I have to put out a PSA about how I get annoyed at unsolicited pitches from people who are one-sided and transactional. These people are takers. I've got no time for takers.
Instead, pay it forward. Give away your ideas in free articles and videos. Mentor people. Create an open source project. Be a friend and a human without asking anything in return. Be a giver.
In this episode, I sit down with Wendy Turner-Williams, a distinguished tech leader and executive with a deep history at companies like Microsoft and Salesforce. She's of the original minds behind what became Azure Data Factory, among other foundational tech.
In this wide-ranging conversation, Wendy charts the trajectory from the early days of the Internet to the current AI-driven hype cycle and looming crisis. She explains how these tools of innovation are now being turned against the workforce and why this technological revolution is fundamentally more disruptive than anything that has come before. This episode is a candid, unfiltered discussion about the real-world impact of AI on jobs, the economy, and our collective future, and a call for leaders to act before it's too late.
Timestamps:
00:22 - Catching up: The tough job market and writing new books.
05:49 - Wendy's impressive career history at Microsoft, Salesforce, and Tableau.
06:17 - The origin story of Azure Data Factory and other foundational projects at Microsoft.
09:18 - A personal story about the challenges of being a woman in Big Tech in the early days.
13:02 - A look back at a favorite early-career project: Digitizing physical maps with nascent GPS technology in 2001.
18:11 - The state of the tech industry: "Tech is cannibalizing itself because of AI."
20:31 - The massive, impending shock to the job market and why AI is different from previous industrial revolutions.
27:26 - Why the "human in the loop" is a temporary and misleading solution.
29:55 - Breaking down the numbers: The staggering quantity of white-collar jobs projected to be eliminated.
36:37 - Why leaders are failing to act and conversations are happening behind closed doors without solutions.
38:25 - Discussing potential solutions: Should companies have quotas for their human workforce?
45:21 - The need for "truth tellers" and leaders who are willing to question the current path and drive human-centric transformation.
53:15 - The grim reality for recent graduates with computer science degrees who can't find jobs.
56:22 - The risk of IP hoarding and engineers deliberately crippling systems to protect their jobs.
01:00:20 - Final thoughts: Are we waiting for a "let them eat cake" moment before we see real change?
Matt Housley joins me for our monthly chat. This time, we discuss my article The Pedantic Layer, fraud, and more.
There are very few people like Stephen Brobst, a legendary tech CTO and "certified data geek," Stephen shares his incredible journey, from his early days in computational physics and building real-time trading systems on Wall Street to becoming the CTO for Teradata and now Ab Initio Software. Stephen provides a masterclass on the evolution of data architecture, tracing the macro trends from early decision support systems to "active data warehousing" and the rise of AI/ML (formerly known as data mining). He dives deep into why metadata-driven architecture is critical for the future and how AI, large language models, and real-time sensor technology will fundamentally reshape industries and eliminate the dashboard as we know it. We also chat about something way cooler, as Stephen discusses his three passions: travel, music, and teaching. He reveals his personal rule of never staying in the same city for more than five consecutive days since 1993 and how he manages a life of constant motion. From his early days DJing punk rock and seeing the Sex Pistols' last concert to his minimalist travel philosophy and ever-growing bucket list, Stephen offers a unique perspective on living a life rich with experience over material possessions. Finally, he offers invaluable advice for the next generation on navigating careers in an AI-driven world and living life to the fullest.
The AI hype cycle seems to be calming down a bit, and this is awesome. Humanity always craves another hype cycle, so I give some ideas on what I think is next after AI.
Ever wonder how companies in the crowded data and AI space build powerful alliances to drive revenue and growth? In this episode, I sit down with Eleanor Thompson, a partnerships expert based in London and founder of a successful partnerships consultancy. Drawing from her experience running the partner program at Fivetran during its hyper-growth phase, Eleanor shares the essential strategies for building a successful partnership ecosystem from the ground up.
We also also discuss the mental fortitude required for entrepreneurship, drawing surprising parallels between running a business and competing in high-intensity fitness events like Hyrox.
Tune in to learn:
The fundamental reasons why partnerships are critical for expanding your reach and generating revenue.
When is the right time for a startup to focus on partnerships (hint: it's not day one).
Eleanor's "4A" framework (Alignment, Ability, Audience, Accountability) for identifying the perfect partner.
The key roles, from Partner Sales Engineers to Partner Ops, needed to build a successful partnerships team.
Red flags to watch for when a potential partner is more focused on margin than customer value.
How AI can be used to identify ideal partners and even predict their future success.
Find Eleanor Thompson online:
LinkedIn: Eleanor Thompson
Website: https://branchworks.io
#BusinessPartnerships #DataEngineering #AI #Entrepreneurship #Tech #Startup #GoToMarket
Timestamps
00:49 - Who is Eleanor Thompson?
02:25 - Why Do Business Partnerships Exist?
05:26 - When is the Right Time for a Company to Start Building Partnerships?
06:40 - The 4A Framework for Defining Your Ideal Partner Profile
08:20 - Joe's Experience Partnering with Big Tech
12:33 - How to Structure a Partnerships Team in a Growing Tech Company
20:49 - What is Partner Operations and Why is It a Critical Hire?
22:30 - The Importance of Trust and Referral Fees
25:15 - Eleanor's Journey as an "Accidental CEO"
30:10 - The Mental Fitness and Resilience Required to Be a Founder
41:20 - How to Use AI in Your Partnership Strategy
45:00 - How to Spot a Good Partner on the Very First Call
46:33 - Red Flags to Watch For in Potential Partners
51:35 - How Fitness and Hyrox Competitions Fuel Business Success
59:45 - Where to Find Eleanor Thompson
Is reality setting in for the AI bubble? Who the hell knows. This is definitely the most bipolar bubble I've ever seen, making the dotcom bubble look downright tame.
In this rant, I discuss why AI is in its televangelist moment, and why a reality check is necessary to keep real progress on AI on track.
Our TOPPODCAST Picks 
