Data River Deep Dive

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode of Data River Deep Dive, Nagim is joined by Yaobang Sheng, Director of Digital Recruitment and Innovation at Akkodis Germany, part of the Adecco Group. Yaobang spent 15 years in IT and engineering recruitment before moving into AI transformation, and he now builds AI consulting capacity for external clients. He comes at AI from the workforce side rather than the engineering side, and that changes almost every answer he gives. The conversation starts with a question most organisations skip. A job description used to take two hours and now takes five minutes. So what happens to the other ninety five minutes? From there we get into why companies buy the model before they fix the process and the data, why cost cutting has become the only AI story anyone tells, and why Yaobang refuses the "data is the new oil" line and calls data a labour layer instead. We talk about what a CV is worth once every candidate writes theirs with AI, why recruiters are going back to in person interviews, and what happens to the company that replaces its entire entry level cohort and then discovers there is nobody left to promote. In the second half Yaobang lays out his three layers of execution, augmented work, and decision and control, and gives us a term worth stealing: FTA, the full time agent. If you lead a team, you will soon be managing both FTE and FTA. He also names the three skills he would hire for now, and none of them are technical. In between there is Draft Night, the AI edition. Six real AI decisions, called as an airball, a mid range, or a slam dunk. Michael Jordan is involved. So is a CDO who probably should be sacked. Whether you lead a data function, sit in HR, or are simply trying to work out what to do with the time AI keeps handing back to you, there is something here to take away.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode of Data River Deep Dive, we sit down with Dr. Joe Perez, Senior Systems Specialist at the North Carolina Department of Health and Human Services, author of From Numbers to Narrative, and one of the most recognisable voices in data storytelling worldwide. Four times on a Times Square billboard, speaking engagements across more than twenty countries, and thirty-three years in state government that he could have left for the private sector at any point and never did. Recorded in London during Big Data LDN, hours after he landed on an overnight flight. Joe does not explain storytelling in this conversation. He performs it. Every answer arrives with a metaphor attached, and by the end you understand exactly why audiences remember what he says. We cover his LIFE framework and which of the four principles organisations consistently skip. The difference between relevance and resonance, and why the number in your presentation matters far less than you think. Where simplification becomes oversimplification and quietly introduces bias into your data story. And what forty years of teaching taught him about attention spans that most data professionals still have not learned. We also play a music-themed game built around his passion for the piano, which produces a failed visualisation described as a pig stuck by a thorn, a career summarised as the Moonlight Sonata, and a story about nearly walking into a lake at 9,428 feet while filming.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode of Data River Deep Dive, we sit down with Tiankai Feng, Director of Data and AI Strategy at Thoughtworks. He spent over six years at Adidas building the practitioner experience that became two books: Humanizing Data Strategy and Humanizing AI Strategy. He did not write the frameworks first. He lived the failures first. That sequence is the credential. Since joining Thoughtworks, Tiankai has been working with organisations worldwide on the question that sits underneath every stalled transformation: why do people not change even when the data is right? His 5C Framework has become a reference point for leaders who want an honest answer to that question rather than a maturity model. This conversation does not stay comfortable. We cover why trust in a team and trust in the data it produces are the same variable. Why creativity is the most misread word in any data strategy. Whether AI can ever imitate not just what you produce but how you actually think. And why the data community needs to talk more honestly about the cultural assumptions baked into the rooms and frameworks it builds. We also built a custom Nintendo game for Tiankai, and it diagnosed the ownership problem better than most governance workshops do.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode of Data River Deep Dive, we sit down with Rosanne Werner, CEO of XcelerateIQ and one of the most distinctive voices on data culture change in Europe. Rosanne spent over a decade at Coca-Cola Europe Pacific, where she built the "Data in Action" programme that mobilised more than 2,000 change catalysts across functions and territories. Since founding XcelerateIQ, she has been helping organisations worldwide close the gap between data investment and actual behavioural change. This conversation cuts through the noise. Rosanne brings neuroscience, organisational psychology, and hard-won practitioner experience to bear on the questions that matter most: Why do data literacy programmes fail? What does psychological safety have to do with dashboards? How do you measure something as invisible as trust? And why do data and AI teams have to walk the same talk they ask of everyone else? We also play a cycling-themed game — Sprint or Endurance — that reveals more about data and AI transformation than most strategy decks ever do.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode, we are joined by Jovita Tam, dual-qualified attorney (England & Wales and New York), Data and AI advisor, board advisor, and one of the most distinctive voices at the intersection of governance, transformation, and responsible AI. Jovita's career spans engineering, law, investment banking, and strategic advisory. A rare combination that gives her an unusually clear view of where organisations deceive themselves, and why. Her thinking does not sit comfortably inside a single discipline, and neither does this conversation. Together, we examine the blind spots that governance frameworks quietly create, the difference between transformation as a project and transformation as a condition of organisational life, and why AND-thinking — holding two realities at once — may be the most underrated leadership capability of our time. We then turn the lens to Generative AI: not as a tool, but as a mirror. One that reflects not only what organisations have built, but what they have been avoiding. Between these deeper reflections, we step onto the mat for our Deep Dive Game: "Data Yoga Flow." Jovita is not only a governance expert and attorney — she is also a yoga instructor. So we put both to work. A data or AI scenario is named, and Jovita calls the pose. It turns out that the vocabulary of yoga and the pathology of organisations have more in common than either would like to admit. This episode is for anyone who has ever suspected that the real transformation was not in the roadmap, but in the willingness to see what the roadmap was built to avoid.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode, we’re joined by Ole Olesen-Bagneux, Chief Evangelist at Actian and O’Reilly author of Fundamentals of Metadata Management (2025) and The Enterprise Data Catalog (2023). Ole’s career bridges library science, enterprise architecture, and modern data platforms, and his insights remind us that metadata is much more than just technical plumbing — it is how organisations define meaning. Together, we explore what it really means to bring meaning and context back into metadata. From the centuries-old wisdom of librarianship to the messy reality of enterprise data catalogs, Ole helps us see how the logic of the library still holds lessons for today’s objectives in data and tech. We discuss why so many metadata systems feel like catalogs abandoned halfway through, and how we can return to designing for discovery. Between these deeper reflections, we lighten the flow with our Deep Dive Game: “The Metadata Bake-Off: Rise It or Slice It?”. A playful yet revealing round where sourdough metaphors meet metadata decisions. Because, as we discovered, data and dough have something in common: both rise only with time, care, and structure. This episode is for anyone who wants to rediscover the meaningful side of information — the quiet art of giving structure to understanding.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode, we’re joined by Frederik Jelden, Global Strategic Account Technology Strategist at Microsoft, whose career bridges enterprise tech strategy, AI governance, and startup grit—yes, including a Vietnamese street food truck. We unpack how organizations can realistically scale GenAI, why governance is still the make-or-break factor, and what it takes to align strategy, tooling, and real adoption in large, complex environments. Frederik shares insights from working with C-level leaders across industries, the challenges of aligning AI ambition with enterprise readiness, and how data infrastructure plays a hidden but decisive role in getting results. We also explore the underestimated cultural layer of AI transformation—who owns what, and why that still matters more than the tooling. If you’re tired of surface-level AI hype and want to hear what happens behind closed doors at large organizations, this one’s for you.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode, we sit down with Dr. Tim Wiegels to explore what it really takes to lead modern data teams—from solo contributor to scaling 100+ people across functions. With experience spanning startups, platforms, and leadership in high-growth environments, Tim offers a raw and practical perspective on navigating KPIs, explaining value, and fighting for relevance inside the business. We dive into the myths of “data-driven” success, the difference between metrics and real performance, and how data leaders often find themselves stuck between business expectations and political crossfire. Expect honest reflection, sharp takes, and a few bold truths on what it really means to justify your existence as a data team.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode, we sit down with Dr. Tim Wiegels to explore what it really takes to lead modern data teams—from solo contributor to scaling 100+ people across functions. With experience spanning startups, platforms, and leadership in high-growth environments, Tim offers a raw and practical perspective on navigating KPIs, explaining value, and fighting for relevance inside the business. We dive into the myths of “data-driven” success, the difference between metrics and real performance, and how data leaders often find themselves stuck between business expectations and political crossfire. Expect honest reflection, sharp takes, and a few bold truths on what it really means to justify your existence as a data team.

Hey! I'd love to hear your thoughts, send me a voice note.

In this episode, we sit down with Martin Treder to unpack the real-world evolution of the Chief Data Officer role. From its misunderstood beginnings to its growing influence in business strategy. Drawing on decades of experience across logistics and healthcare, Martin shares what it truly means to treat data as an asset, why data culture is still the toughest nut to crack and how CDOs are often tasked with much more than just analytics. We explore his transition from operational roles into enterprise data leadership, the critical role of SAP ecosystems, and ifthe CDO might be closer to a modern COO than we think. Expect sharp insights, honest reflections and a fresh take on what CDOs really do and don’t do.