AI Stories

Discover how Demetrios Brinkmann transitioned from teaching to founding the MLOps Community. Explore the role of MLOps in the ML lifecycle and the debate between specialized vs generalizing language models. Learn about the business implications of relying on OpenAI and the activities of the MLOps Community, including live events, podcasts, and new courses on generative AI.

In this discussion, Noah Gift, MLOps leader and executive in residence at Duke University, shares insights from his 30 years of experience, including building data pipelines in the film industry. He emphasizes the crucial role of MLOps and the software engineering skills essential for data scientists. Noah contrasts Python and Rust, advocating for flexibility in choosing tools. He delves into the differences between MLOps and LLMOps, discussing security concerns and the future of deployment strategies, making a compelling case for adapting to the tech landscape.

Our guest today is Marianne Ducournau, Head of Data Science at Qonto and ex Data Scientist at Amazon and Uber.In our conversation, we first discuss Marianne's first job in Data Science working in the public sector and managing a 10-15 people team. Marianne then talks about her experience at Uber and shares various projects that she worked on. We dive into price elasticity modelling and financial forecasting where her team built thousands of model to forecast financial metrics in multiple cities.  Marianne finally explains her current role as the Head of Data Science at Qonto and gives advice on how to progress in Big Techs and in your career. If you enjoyed the episode, please leave a 5 star review and subscribe to the AI Stories Youtube channel.Link to Train in Data courses (use the code AISTORIES to get a 10% discount): https://www.trainindata.com/courses?affcode=1218302_5n7krabaFollow Marianne on LinkedIn: https://www.linkedin.com/in/mborzic/Follow Neil on LinkedIn: https://www.linkedin.com/in/leiserneil/  ————(00:00) - Introduction(02:12) - Marianne's Journey Into Data Science(05:05) - Managing A 10-15 People Team In Her First Job(10:02) - Pros And Cons Of Working In The Public Sector(16:51) - Transition From The Public Sector To Uber(22:25) - Price Elasticity Modelling(35:42) - Building 1000+ Models For Financial Forecasting(42:10) - Progressing In Big Techs(45:01) - What Is Qonto And Marianne's Role There?(48:08) - Understanding Qonto's Product(49:29) - Building A Team As Head Of Data Science(54:37) - Impact Estimation(01:02:52) - Marianne's Advice For Career Progression

Christof Henkel, Senior Deep Learning Data Scientist at NVIDIA and world number 1 on Kaggle, discusses his journey on Kaggle, the role of mathematics in Data Science careers, winning competitions, reducing model overfitting, and advice on progressing in your career.

Davis Blalock discusses his journey from academia to joining Mosaic ML, the startup's focus on efficient neural network training, evolution to large language models, and the $1.3 billion acquisition by Databricks. Topics include enhancing ML algorithms' efficiency, importance of efficiency, and learnings from working in a startup.

Our guest today is Kellin Pelrine, Research Scientist at FAR AI and Doctoral Researcher at the Quebec Artificial Intelligence Institute (MILA). In our conversation, Kellin first explains how he defeated a superhuman Go-playing AI engine named KataGo 14 games to 1. We talk about KataGo’s weaknesses and discuss how Kellin managed to identify them using Reinforcement Learning. In the second part of the episode, we dive into Kellin’s research on building practical AI systems. We dig into his work on misinformation detection and political polarisation and discuss why building stronger models isn’t always enough to get real world impact. If you enjoyed the episode, please leave a 5 star review and subscribe to the AI Stories Youtube channel.Follow Kellin on LinkedIn: https://www.linkedin.com/in/kellin-pelrine/Follow Neil on LinkedIn: https://www.linkedin.com/in/leiserneil/  ————(00:00) - Intro(01:54) - How Kellin got into the field(03:23) - The game of Go (06:10) - Lee Sedol vs AlphaGo(11:42) - How Kellin defeated KataGo 14 -1(26:24) - Using AI to detect KataGo’s weaknesses (37:07) - Kellin’s research on building practical AI systems(43:10) - Misinformation detection (49:22) - Political polarisation(54:39) - ML in Academia vs in Industry(1:06:03) - Career Advice

Our guest today is Chanuki Seresinhe, head of Data Science at Zoopla,  a company which provides millions of users with access to properties for sale and for rent. In our conversation, we first talk about Chanuki’s PhD where she used machine learning to identify relationships between beautiful places and happiness. We then dive into Data Science at Zoopla and talk about Generative AI and other exciting projects that Chanuki is currently working on. Throughout the episode, Chanuki shares great insights on why ML projects fail, the importance of good metrics, switching companies and how to progress in your career. If you enjoyed the episode, please leave a 5 star review and subscribe to the AI Stories Youtube channel.Follow Chanuki on LinkedIn: https://www.linkedin.com/in/chanukiseresinhe/Follow Neil on LinkedIn: https://www.linkedin.com/in/leiserneil/  ————(00:00) : Intro(01:23) : How Chanuki got into the field(04:58) : AI to better understand happiness(16:37) : Generative AI (21:26) : Generative AI vs supervised learning(24:47) : Data Science at Zoopla(31:46) : The importance of good metrics(35:33) : Dealing with outliers(39:41) : Why ML projects fail(46:30) : Switching companies(48:42) : Bias (54:47) : Career advice

Rémi Ounadjela shares his experience working on shipment optimization at Amazon and data science for risk and safety at TikTok. He discusses the differences between working at TikTok, Google, and Amazon. Rémi also gives advice on how to land a data science job in big tech and common mistakes to avoid.

Barr Moses, Co-Founder and CEO of Monte Carlo, shares her inspiring journey from data strategy at Bain to building a pioneering data observability platform. She discusses her unique experiences in the Israeli Army and highlights the importance of teamwork and feedback in data science. The conversation dives into the critical need for reliable data, illustrated by a staggering $100 million loss due to a data error. Barr also explains the five pillars of data observability and emphasizes how data accuracy is vital for trust, especially in an AI-driven future.

Our guest today is Parul Pandey, Principal Data Scientist at H2O.ai, Kaggle Grandmaster (notebooks) & book author of “Machine Learning for High Risk Applications”. In our conversation, we first dig into Kaggle. Parul explains how she became a Grandmaster, shares tips about data analysis and discusses the pros of learning on Kaggle. The second part of the episode is around machine learning for high risk applications. We talk about the risks of using AI to make decisions, talk about interpretable algorithms and give advice on how to deploy robust models. If you enjoyed the episode, please leave a 5 star review and subscribe to the AI Stories Youtube channel.Machine Learning for High Risk Applications: https://www.oreilly.com/library/view/machine-learning-for/9781098102425/Kaggle notebook (gold medal), "Geek Girls Rising : Myth or Reality!": https://www.kaggle.com/code/parulpandey/geek-girls-rising-myth-or-reality/notebookBlog post on Explainable Boosting Machines: https://towardsdatascience.com/the-explainable-boosting-machine-f24152509ebbFollow Parul on LinkedIn: https://www.linkedin.com/in/parulpandeyindia/Follow Neil on LinkedIn: https://www.linkedin.com/in/leiserneil/  ————(00:00) : Intro(01:57) : How Parul got into AI(07:20) : Kaggle & Becoming a Grandmaster(13:52) : Advice for good data analysis(20:27) : Pros of learning on Kaggle(24:28) : ML for high risk applications(49:20) : Interpretable algorithms (55:28) : Deploying robust models(01:01:22) : Future of AI(01:05:47) : Career advice