Episodes

  • How Data Scientists Use Reinforcement Learning for Dynamic Pricing

    How Data Scientists Use Reinforcement Learning for Dynamic Pricing
    Jun 15 2026
    In this episode of The Data Science Podcast, Lucas and Luna explore how reinforcement learning (RL) is transforming dynamic pricing strategies. Using the example of a major ride-hailing company, they break down how RL algorithms learn to set prices in real time by balancing exploration (testing new price points) and exploitation (using known optimal prices). Lucas explains the core RL concepts of state, action, reward, and the epsilon-greedy algorithm. Luna digs into the practical trade-offs: how often should a model explore versus exploit, and why the reward function must account for long-term customer retention, not just immediate revenue. The conversation also touches on how RL differs from A/B testing in dynamic pricing, the role of simulation environments for training, and ethical considerations around price discrimination. Listeners will walk away with a concrete understanding of RL-based pricing mechanics and a mental model to evaluate pricing algorithms they encounter daily. #ReinforcementLearning #DynamicPricing #DataScience #MachineLearning #RL #PricingStrategy #RideHailing #ExplorationExploitation #EpsilonGreedy #RewardFunction #AIBusiness #Technology #TechPodcast #DataDriven #FexingoBusiness #BusinessPodcast #DataSciencePodcast #Fexingo Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    9 mins
  • How Data Scientists Build Churn Prediction Models That Actually Work

    How Data Scientists Build Churn Prediction Models That Actually Work
    Jun 14 2026
    Churn prediction is one of the most common — and most frustrating — problems data scientists face. In this episode, Lucas and Luna dig into why many churn models fail in production and what separates the ones that actually reduce customer loss. They walk through a concrete example from a mid-size telecom company that cut churn by 14 percent in six months by focusing on the right features and the right deployment strategy. Along the way, they discuss the trap of over-relying on recency features, why boosting models often beat neural nets on tabular churn data, and how to build a simple intervention framework that turns predictions into action. If you've ever built a churn model that looked great in the notebook but went nowhere in production, this one's for you. #ChurnPrediction #DataScience #MachineLearning #CustomerAnalytics #FeatureEngineering #XGBoost #Telecom #ModelDeployment #PrecisionRecall #CustomerRetention #LifetimeValue #TinyML #ProductionML #BusinessAnalytics #Technology #FexingoBusiness #BusinessPodcast #DataDriven Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    6 mins
  • How Data Scientists Use Active Learning to Label Smarter

    How Data Scientists Use Active Learning to Label Smarter
    Jun 14 2026
    In this milestone 50th episode, Lucas and Luna explore active learning — a machine learning paradigm where the model itself chooses which data points to label, dramatically reducing manual annotation costs. They break down the core idea using a concrete example: training a fraud detection model for a payment processor processing 10 million transactions per day. Lucas explains uncertainty sampling, query-by-committee, and the 'exploration vs. exploitation' trade-off. Luna raises the practical challenge of label noise and how to handle it. They also discuss when active learning fails — like when the unlabeled pool doesn't represent real-world distribution. The conversation ties back to the broader theme: getting more value from fewer labels, a critical skill for any data scientist working with limited annotation budgets. #ActiveLearning #MachineLearning #DataScience #UncertaintySampling #QueryByCommittee #FraudDetection #Labeling #Annotation #SemiSupervisedLearning #ExplorationVsExploitation #ModelTraining #DataEfficiency #MLStrategy #Technology #Podcast #FexingoBusiness #BusinessPodcast #DataSciencePodcast Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    10 mins
  • How Data Scientists Use Distributed Computing for Massive Datasets

    How Data Scientists Use Distributed Computing for Massive Datasets
    Jun 13 2026
    When your dataset outgrows a single machine, what do you do? In this episode, Lucas and Luna explore how data scientists use distributed computing frameworks like Apache Spark and Dask to process terabytes of data without crashing their laptops. They break down the key concept of data partitioning, explain why MapReduce is still relevant, and walk through a real example of how a mid-sized e-commerce company reorganized its log-processing pipeline to cut runtime from 14 hours to 47 minutes. Lucas shares a cautionary tale about shuffling bottlenecks that can ruin a cluster's performance, and Luna asks the practical question every team faces: when does it make sense to move from a single-node pandas workflow to a distributed system? They also discuss managed services like Databricks and AWS EMR versus rolling your own cluster. No prior distributed systems experience required — just a curiosity about what happens when data gets too big for a spreadsheet. #DataScience #DistributedComputing #ApacheSpark #Dask #MapReduce #BigData #DataEngineering #DataPartitioning #Shuffling #Databricks #AWSEmr #Pandas #Tech #Technology #FexingoBusiness #BusinessPodcast #Podcast #DataPodcast Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    8 mins
  • How Data Scientists Are Using TinyML on Edge Devices

    How Data Scientists Are Using TinyML on Edge Devices
    Jun 13 2026
    This episode of The Data Science Podcast dives into TinyML—the practice of running machine learning models on microcontrollers and edge devices with limited power and memory. Lucas and Luna explore a specific case: a smart agriculture startup that used a TensorFlow Lite model on an Arduino board to detect crop disease in real time, processing images locally without sending data to the cloud. They discuss the trade-offs: model compression techniques like pruning and quantization, the challenge of balancing accuracy against a 256KB memory budget, and why TinyML is gaining traction in industries from manufacturing to healthcare. The hosts also touch on the broader movement toward on-device AI, privacy benefits, and the emerging toolchain from Google's TensorFlow Lite Micro to ARM's Ethos-U55. If you've ever wondered how data scientists shrink neural networks to run on a battery-powered sensor, this episode is for you. #TinyML #EdgeAI #MachineLearning #DataScience #EmbeddedSystems #TensorFlowLite #SmartAgriculture #ModelCompression #Pruning #Quantization #OnDeviceAI #IoT #Microcontrollers #PrivacyPreservingML #Technology #FexingoBusiness #BusinessPodcast #DataDriven Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    8 mins
  • How Data Scientists Use NLP to Detect Misinformation

    How Data Scientists Use NLP to Detect Misinformation
    Jun 12 2026
    In this episode, Lucas and Luna dive into the growing role of natural language processing in detecting online misinformation. They explore a specific case study: how DataGPT, a startup, built a fact-checking bot that flags false claims in real-time on social media. Lucas breaks down the technical stack — transformer models like BERT fine-tuned on fact-checking datasets, with a focus on stance detection and claim verification. Luna questions the reliability of automated fact-checking and raises the issue of adversarial attacks on NLP models. They discuss the 2024 US election as a major test case, where the bot achieved 83% accuracy. The episode also touches on the ethical trade-offs: is automated fact-checking effective or does it risk censorship? #NaturalLanguageProcessing #Misinformation #FactChecking #DataGPT #BERT #TransformerModels #StanceDetection #ClaimVerification #2024Election #AIEthics #TechEthics #AdversarialAttacks #SocialMedia #Technology #FexingoBusiness #BusinessPodcast #DataScience #MachineLearning Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    10 mins
  • Data Scientists Are Using Graph Neural Networks for Fraud Detection

    Data Scientists Are Using Graph Neural Networks for Fraud Detection
    Jun 12 2026
    Episode 46 of The Data Science Podcast with Fexingo explores how graph neural networks (GNNs) are transforming fraud detection in financial services. Lucas and Luna break down a real case: how a major European bank used GNNs to catch a money-laundering ring that rule-based systems missed. They discuss why traditional fraud models fail with relational data, how GNNs represent transactions as a graph of accounts and connections, and what it means for data scientists building production pipelines. Topics include message-passing layers, inductive vs transductive learning, and the challenge of class imbalance in fraud datasets. The hosts also touch on explainability trade-offs and whether GNNs are ready for real-time scoring. If you're a data scientist curious about graph-based machine learning beyond social network recommendations, this episode gives you a concrete, business-driven example to learn from. #GraphNeuralNetworks #FraudDetection #MachineLearning #DataScience #FinancialServices #AntiMoneyLaundering #GNN #GraphML #MessagePassing #InductiveLearning #ClassImbalance #Explainability #ProductionML #Technology #DataSciencePodcast #FexingoBusiness #BusinessPodcast #ModelDeployment Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    8 mins
  • How Data Scientists Use Counterfactual Explanations to Build Trust

    How Data Scientists Use Counterfactual Explanations to Build Trust
    Jun 11 2026
    Episode 45 of The Data Science Podcast explores the emerging practice of counterfactual explanations — 'what-if' scenarios that help end users understand why a machine learning model made a particular decision. Lucas and Luna walk through a concrete example from the lending industry: a small business owner named Maria whose loan application was denied by an automated risk model. Instead of a black-box rejection, a counterfactual explanation tells her: 'If your annual revenue were $150,000 instead of $100,000, your application would have been approved.' The hosts discuss how companies like JPMorgan and Zest AI are piloting these techniques to comply with regulatory pressure and improve customer trust, and they weigh the trade-offs between fidelity and simplicity. They also touch on the computational cost of generating counterfactuals at scale and the risk of exposing sensitive model boundaries. This episode is anchored to the current regulatory landscape as of June 2026, with references to the EU's AI Act and the FTC's guidance on algorithmic fairness. #CounterfactualExplanations #ExplainableAI #MachineLearning #DataScience #LendingAlgorithms #ModelInterpretability #AITrust #RegulatoryCompliance #EUAIAct #FTC #JPMorgan #ZestAI #SmallBusinessLoans #WhatIfAnalysis #BlackBoxModels #AlgorithmicFairness #FeatureImportance #FexingoBusiness Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    9 mins