Winning With Data

Michael Parekh is a former partner and managing director at Goldman Sachs.  Today, he serves those same roles at MKP Capital Management, focusing on AI and the 4th Tech wave.  Michael was using the internet in the 1980s, and has closely watched the evolution of all things digital.  He regularly shares his insights with the world on his personal Substack.In this episode, Jake and Michael discuss Michael's theory of the "AI value stack", why we are at "the beginning of the beginning", what's next after the chat bot, and the ongoing race to build the first AI market moat.

10/7/24 • 32:30

Tim Chen is a managing partner at Essence Venture Capital and part-time advisor to Bessemer Venture Partners.  He was a principal engineer at Grid.ai, which created a high-level interface for the Pytorch deep learning network.  Cloudera acquired Tim's machine-learning-based startup, Hyperpilot, in 2018.In this conversation, Tim and Jake drill down on what A.I. businesses and products venture capital is really looking to fund, how to tell the hype from real value, and how to think about a product's long vs. short term disruption of a market.

8/23/24 • 29:09

Ben Rosencranz spent several years as a strategy and analytics manager with E15 before signing on with the Chicago Blackhawks.  He is currently using advanced analytics and technology to drive decision making for the team, on and off the ice.In this conversation, Jake and Ben discuss how he chooses which data problems to tackle, the exciting new changes data is bringing to fan experience and engagement, and how to deal with unrealistic expectations for AI.

7/26/24 • 21:22

Ken Jee is the head of data science for Scouts Consulting Group.  In addition to advising senior sports executives, Ken also hosts his own podcast, He also hosts Ken’s Nearest Neighbors podcast, which he describes as "a quest to compile the most complete study on sports performance ever created."In this episode, Ken discusses what successful data-driven organizations are doing right, starting with a problem vs. starting with a data set, and the tricky question of who gets access to data.

7/23/24 • 19:30

Andrew Ehrenberg is the Head of Business Development at WagerWire.  He has ownership stakes in four sports organization and full partnership in Eberg Capital, where he's making big investments in "disruptive sports wagering applications and the commercialization of alternative sports."  He recently found himself in the Kentucky Derby winners circle after syndicating 25% of the winning colt, Mage, to a community of 400 investors.In this interview, Jake and Alec pick Andrew's brain about what he learned as analytics product manager at DataRobot, how sophisticated private equity firms are professionalizing sports analytics, and why AI generalists are going to be pushed out of the market.

7/12/24 • 20:07

Ben Alamar became Director of Sports Analytics for ESPN a decade ago after the publication of his pioneering book, "Sports Analytics: A Guide for Coaches, Managers, and Other Decision Makers" in 2013.  Since then, Ben has built and managed analytics teams at StubHub and the Cleveland Cavaliers, and is currently working as a data science manager for Apple.  In Ben's second appearance on the show, we talk about the insights found in the newest edition of his book, which updates all discussions for the rapid technological developments of the last decade, adding strategies that focus on using data on the business side of sports.

6/25/24 • 25:29

Hadi Sotudeh ⚽📈 is a doctoral student at ETH Zurich studying team tactical formations such as 4-4-2 and 3-5-2 using tracking data.He has previously worked as a football data science consultant for the Performance & Technology department of KNVB (Royal Dutch Football Association) and placed in the top 1% of contestants at the Paris Saint-Germain’s sports analytics challenge.In this episode, Hadi discusses the specifics of his work at ETH Zürich, when it makes sense to buy rather than build data solutions, the biggest problems teams can solve using tracking data, and his football championship predictions.

6/10/24 • 17:26

Varuna De Silva leads the Machine Intelligence Lab at Loughborough University, applying artificial intelligence to engineering problems.  This work has included the integration of A.I. and sports data analytics to provide insights for national and international teams and clubs.In this episode, Varuna, Alec, and fellow Loughborough alum Jake discuss how universities teach students in the rapidly changing world of A.I., how to see through hype in the field, why sports insights must make their way to players, and his analysis of the build vs. buy debate.

6/7/24 • 18:46

Jessica Gelman has been an influential player in sports data ever since co-founding the MIT Sloan Sports Analytics Conference all the way back in 2008. She is currently the CEO of Kraft Analytics Group and minority partner in the Utah Royals of the National Women's Soccer League. Jessica has designed analytics to build high-performing teams, but also to optimize ticket sales, merchandising and customer insights.In this episode, Jessica shares how her struggles as a slow, short point guard in college basketball led her to a lifelong passion for sports data, the three data pain points she sees in organizations most often, and the importance of stepping outside the algorithms every now and then.

5/8/24 • 33:20

Keith Goldner is the Vice President of Data Science, Sports Modeling & Innovation at FanDuel, one of the largest sports betting platforms in the U.S. Keith has over a decade of experience as a consultant with numerous national franchises, as well as contributing analysis to ESPN, Sports Illustrated, and the Wall Street Journal.In this episode, Keith explains why it’s crucial to eliminate single-person dependencies, the importance of listening to the opinions of stakeholders and the public (regardless of what your model says), and the necessity of communicating effectively with your organization.

5/1/24 • 20:42

Zack Scott is the founder and CEO of Four Rings Sports Solutions and adviser for Gemini Sports Analytics. Over the past 20 years, Zack’s data-driven leadership has directly contributed to four Boston Red Sox world championships. He spent one year as General Manager of the New York Mets, and he has advised the Pittsburgh Penguins, Detroit Tigers, and 2023 World Series champion Texas Rangers.In this episode, Zack reveals the common themes among winning sports clubs, including how to explain models with transparency and honesty, the balance between data and game expertise, and the importance of healthy conflict in decision making.

4/26/24 • 28:11

Suds Gopaladesikan is the director of football intelligence for Italy's Atalanta BC. Mathieu Lacome is the head performance and analytics officer with the Parma Calcio football club in Italy, where he develops on-pitch and scouting strategies. Both teams are greatly outperforming their squad costs and operating in a sustainable manner. Together, Suds and Mathieu bring incredible perspective on how data analytics can successfully play out on the pitch.In this episode, they talk about the key components of that success, what they’ve learned about data roles in European soccer clubs, how to drive effective organizational change, and why touchpoints within the organization should be a key performance indicator. 

4/22/24 • 24:45

Matthew Lynley has been a journalist covering the tech industry for more than a decade. Four years of this career have been spent covering analytics and data science - tools which he personally used. He is the founder of the Supervised newsletter that aims to help decision-makers sort through the hype and reality of AI tools and the future of big data analytics.In this episode, Matthew goes over security for proprietary data troves, how that data may be sold or rented in the near future, the challenges of unstructured data, how big data could play into drafting, and foundation models vs. tooling and fine tuning.

4/18/24 • 28:04

Seth Stephens-Davidowitz is a data scientist who put his Harvard PhD to work at Google, as well as consulting for numerous Fortune 500 companies.  His first book, Everybody Lies, was a New York Times bestseller and an Economist Book of the Year.  Almost the entirety of his new book, Who Makes the NBA, was written over the course of a single month by leveraging AI data tools.In this conversation, Seth, Jake and Alec discuss some of the incredible insights the book reveals, how AI is replacing coding with critical thought, and what the future looks like for sports data analytics.

4/9/24 • 19:27

Matt Ober is an extremely accomplished data leader turned venture capitalist. After leading Bloomberg's data operations for years, Matt served as chief data scientist for the $20 billion hedge fund Third Point, where he led development and implementation of data and analytics strategies for the entire firm.  He is currently general partner at Social Leverage, an early stage seed investment fund.In this conversation, Matt shares how data practices from the financial world can apply in the world of sports, why AI will become just another part of tech stacks, the key difference between a winning and losing data organization, and what the best analysts will be doing 15 years from now.

4/3/24 • 27:24

Miles Adkins has spent a decade in the data science world, from hedge fund investment models to his current position as A.I. and Machine Learning specialist at Snowflake cloud computing.  He has served in both administrative and customer-facing roles, and is intimately familiar with the successful implementation of data strategies.In this episode, Miles, Jake and Alec discuss what it was like at the beginning of the automation revolution, why speed can be more important than building custom models, why "on prem" data sacrifices too much, and what's causing all the mergers and acquisitions in big data warehousing companies.

3/27/24 • 22:45

Achille de Rauglaudre is a Sports investor and advisor.  Mathieu Lacome is author of the book "Football Intelligence" and the head performance and analytics officer with the Parma Calcio football club in Italy, where he develops on-pitch and scouting strategies.  Together, Achille and Mathieu bring a substantial body of knowledge and experience to the table.In this conversation, we discuss why a business strategy is necessary to effectively implement data strategy, what that looks like for successful teams, the importance of setting realistic expectations and goals, and why communication and visualization of data is so valuable for internal cohesion.

3/22/24 • 34:13

Self-proclaimed data nerd David Spezia served as one of the original Field CTO's at Tableau, was the Principal Sales Engineer at Snowflake during the largest IPO in software history, and is now a Distinguished Architect at Salesforce. His specialties include problem solving with technology, data visualization, and information system design and architecture.In this conversation, David, Jake and Alec discuss the ways successful businesses leverage data technology and platforms, how predictive analytics are evolving, the simple questions David asks to help clients design user experience, and what hiking can teach us about creating those systems.

3/14/24 • 29:21

David Hardoon is an executive with more than two decades of international data science experience in multiple fields.  In the financial sector, he headed up data and AI strategy for the central bank of Singapore and UnionBank of the Phillipines.  In academia, David has served as an Honorary Senior Research Associate with the University College of London for many years.  In industry, he currently serves as CEO of Aboitiz Data Innovation. In this conversation, David, Jake and Alec will discuss "centers of excellence", how they can go wrong, why identifying core needs comes before tech and data, how ownership and use of data insights within organizations might be best distributed, and more.

3/7/24 • 30:36

Jordan Morrow is the founder and owner of Bodhi Data. Since 2019, he’s won 7 awards for his work on data, analytics, and AI. Bodhi Data is aiming to transfer skills in these fields to individuals and organizations that will allow them to make data-driven decisions without becoming advanced data professionals themselves.In this conversation Jordan, Jake, and Alec discuss the incredible pressures that coaches and GMs are under to implement new data and AI solutions, why organizations should resist chasing all the new trends, and what we can learn from the changing fortunes of New Zealand and South African rugby.

2/27/24 • 30:19

Andrew Madson is the Senior Director of Data Analytics at Arizona State University, while also serving as professor at three other universities.  Andrew is a solutions-focused analyst, specializing in machine learning, statistical analysis, and data visualization. His team at ASU is working to bridge their academic research with industry for real-world impact.In 2023, Andrew founded Insights X Design, a data analytics firm that helps businesses with user adoption, machine learning, and targeted A.I. In this conversation, we discuss the failed promise of dashboards and self-serve analytics, why it's becoming more popular to think about data as a product in itself, the trap of over-optimization, and the proper role of LLMs.

2/18/24 • 19:45

Jesse Davis is a computer science professor at KU Leuven, Belgium, where he supervises teams of PhD students researching A.I., data science, and machine learning. Jesse is probably best known for co-developing Valuing Actions by Estimating Probabilities (VAEP) framework in soccer, now widely used throughout the game. He is also cofounder of RunEASI (runeasi.ai), a wearable tech company that integrates A.I to analyze the performance of athletes.In this episode, Jesse, Jake and Hugo dig into the framework that underpins the VAEP's framework, how it can be used to enhance team recruiting, the qualities that make for good models, and the connections between data analytics across different sports.

2/15/24 • 22:01

Day-Yi is a lecturer at Princeton's Center for Statistics & Machine Learning.  Yi has taught college-level math courses, trained corporate clients, and built his own consulting practice focused on data technologies. His client list includes companies such as New York Life, Unilever, Morgan Stanley, and Bank of China. He has also taught at Columbia, where he was selected as a finalist for Columbia’s Presidential Award for Outstanding Teaching in 2021.  Yi is passionate about three things: sports, data and education. In this episode, Yi discusses how he's creating unusual blends of all three. He'll explain why college students are lining up to take his Princeton class on Taylor Swift, and how that connects to a Monte Carlo simulation that has so far predicted playoffs with 100% accuracy. Yi has lots of valuable advice to share - from building a data analytics career to the best Korean restaurant in New York City.

2/8/24 • 21:17

Edem Spio, co-founder, co-host and senior producer of the Africa Business of Sport Podcast and co-founder and director of the Africa Sport Business Group. Spio is also the Media Relations Office at Toppem Sports, and the Public Relations Officer at Yong African Leaders in Sports. In this episode, Spio talks about the integration of AI and data analytics in the production of his podcast, strategies to harness data for enhancing decision-making in Lower League Management (LLMs), Wimbledon's innovative use of virtual reality (VR) and its potential to captivate a broad spectrum of sports fans and a dialogue on the greatest of all time (GOATs) in tennis, rugby, and soccer/football.

1/31/24 • 20:05

Eric Siegel, Ph.D., is a former Columbia University professor, founder of the Predictive Analytics World and Deep Learning World conference series. Siegel authored the bestselling book, "Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die," which is a textbook at university courses worldwide.In this episode, Eric Siegel talks about his new book, The AI Playbook: Mastering the Rare Art of Machine Learning Deployment, and the core principles it lays out for ML implementation. This condensed discussion includes practical knowledge such as: Why most data scientists are too technical for their own good, how UPS saved millions with effective implementation of ML, the organization-wide onboarding required to make lasting changes through analytics and why hype about AGI is sometimes irrelevant in a world where most businesses haven't even learned the core lessons of ML and predictive analytics.

1/30/24 • 25:32

Nancy Hensley is a powerhouse in the sports industry and data-driven technology. After serving as the Chief Digital Officer for IBM's Data and AI division, she revolutionized the AI product offerings at Stats Perform, before going on to co-found women's soccer conglomerate Mercury 13.In this episode, Nancy talks about what it takes to be a part of a data company in the sports world and what makes that industry unique. Nancy also shares her thoughts on how to best integrate AI into product offerings and where the sports analytics world is heading in the near future.

1/28/24 • 18:56

Welcome back to the Winning with Data Podcast! Jepson Taylor has 18 years of experience in machine learning. His career journey began at Intel/Micron, and he later transitioned into a pivotal role at a hedge fund. Jepson then served as Chief Data Scientist at the Sequoia-backed HR startup, HireVue. In 2017, Taylor co-founded Zeff.ai with Gemini advisor David “Gonzo” Gonzalez, which was acquired by DataRobot, where he assumed the role of Chief AI Evangelist. Jepson moved from there to become the Chief AI Strategist at Dataiku in 2022 before leaving to launch his second (still stealth!) startup at the end of 2023. His significant impact in developing innovative solutions, such as video interview predictions and network design integration, demonstrates Taylor's commitment to leveraging AI for transformative outcomes in both professional and personal realms.In this episode, Jepson Taylor speaks about the democratization of data and how that can be achieved across organizations in a way that enables all employees to make better decisions through effective collaboration. They also discuss AI technology and its applications in and impact on business, and how sports teams can use such tech for modern, sustainable decision-making. Taylor also gives invaluable insight into storytelling, intellectual honesty and achieving success in business.

1/19/24 • 34:37

Simba Khadder has a decorated history in tech; founding and acting as CEO of Featureform, formerly known as StreamSQL, founded in 2020. Working up from base level engineering roles Khadder has seen tech at every level, giving him a unique perspective and credibility when it comes to speaking on machine learning, data pipelines, and the business processes surrounding projects in that space. A champion of new era tech startups leveraging data and predictive analysis as a service, Khadder has a keen mind for recognizing what spaces or industries that data scientists can benefit from assistance in breaking down the often overwhelming tidal wave of raw data faced.In this episode of the Winning With Data podcast, Simba Khadder speaks on the importance of feature engineering in the world of machine learning and how it is oftentimes the most important portion of any data science project pipeline. Leveraging his own experience and time spent around the industry, Khadder delves into the emergence of data as a product. Stressing the importance of seeing data as more than something that adds to existing teams but as something that adds value by way of generating insights and optimizing processes. Validating this stance through an explanation on how to quantify how important any given data-centered team is to a company, Khadder doubles down on the importance of end-user experience and how a well resourced data science team can drive that bottom line.

6/21/23 • 25:10

David “Gonzo” González has well over a decade of experience putting AI and ML into production in mission-critical systems. Currently working on a stealth startup in the psychedelics space, he previously co-founded Zeff.ai, acquired in 2020 by DataRobot where then served in an executive role bringing a jobs-to-be-done perspective to focus product strategy. Known as a pioneer of deep-learning-as-a-service and transactionally authored training and inference on multi-modal datasets, Gonzo authored The AI Manifesto for Applied Artificial Intelligence Development and has multiple ML patents.In this episode of the Winning With Data podcast, David “Gonzo” Gonzalez talks about how AI, specifically the release of ChatGPT, is not well-designed for casual users at the moment. He remarks that it's better to think of it as an external data source and find value with it. He also stresses the importance of building an understanding of your book of business to not waste resources. A good organization has a strong data- and insights-driven staff and builds its models based on their unique understanding of the field, without room for people or systems that don't focus on the productization data. Gonzo talks about avoiding and working through mistakes in the analytics field, such as avoiding falling into the trap of starting a project all over when a small mistake is made.

6/9/23 • 27:19

Sergio Santamaria, Director of Business Innovation and Data Solutions at Sports Info Solutions, has experience as a data analyst for the Detroit Pistons, the Phoneix Suns, and the Houston Rockets. He also spent time at Stats Perform and founded his own company, 3D Sport, which focuses on providing equal inclusion for all in the sports industry.In this episode of the Winning With Data podcast, Sergio Santamaria talks with Jake about the challenges of working with qualitative data and the importance of having clean datasets. He discusses how everyone in an organization should build some technical skills to increase overall efficiency and help free up analysts' workflows. Sergio talks about how alignment and consistency from organizational leadership  and true adoption of the principles laid out by each team as important markers success. He remarks that all analytics teams should challenge themselves to make systems and frameworks that incorporate both qualitative and quantitative data and build these systems to last.

6/2/23 • 24:00