Data teams have embraced versioning their data like code, but the practice carries a hidden cost that most teams discover too late. In this episode, Lucas and Luna unpack a 2025 survey that found a third of data teams now version their data — and why 40 percent of them report pipeline failures after a version rollback. They dig into the operational burden of versioning: the backup bloat, the semantic drift, and the governance gray zone where versioned data conflicts with compliance policies. Through the lens of a mid-sized retail analytics team's migration to a lakehouse, they explore practical strategies to keep versioning useful without turning it into a tax on your storage budget or your engineers' time. From time-travel queries to the perils of versioning sensitive data, this episode gives you the questions to ask before you version everything. Listeners leave with a clear-eyed view of when versioning pays off and when it becomes a liability.

#DataVersioning #DataEngineering #DataGovernance #DataStorageCosts #LakehouseArchitecture #TimeTravelQueries #DataPipelines #DataCompliance #DataQuality #BusinessIntelligence #Analytics #DataInfrastructure #TechOps #DataStrategy #BigData #FexingoBusiness #BusinessPodcast #DataLeaders

Keep every episode free: buymeacoffee.com/fexingo

Discover how one of Europe's largest retail banks slashed query latency by 60% and storage costs by 40% by adopting a lakehouse architecture built on Delta Lake and object storage. Lucas and Luna break down what makes the lakehouse pattern different from traditional data warehouses and lakes, including ACID transactions, schema enforcement, and open table formats. They discuss the migration challenges, governance wins, and why this approach is reshaping data engineering. Plus, a look at Apache Iceberg and the open-source ecosystem driving adoption. This episode is perfect for data leaders evaluating a unified analytics platform.

#LakehouseArchitecture #DeltaLake #ApacheIceberg #DataLake #DataWarehouse #OpenTableFormats #ObjectStorage #DataEngineering #Analytics #QueryPerformance #CostOptimization #DataGovernance #EuropeanBank #Spark #DataMigration #BusinessAndTechnology #FexingoBusiness #DataBusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

When a fintech team changed a source column type last month, their data lineage tool automatically identified every downstream dashboard and model that would break — before the change went live. In this episode, Lucas and Luna unpack how automated impact analysis transforms data lineage from a post-mortem artifact into a proactive guardrail. They walk through a real-world example: a mid-sized fintech that saved a 45-minute outage by connecting lineage metadata to their CI/CD pipeline. They discuss the key technical components — open lineage events, column-level tagging, and automated notification hooks — and why this approach is gaining traction as data teams shift left on quality. Plus, a quick mention of how listener support keeps the show ad-free.

#DataLineage #ImpactAnalysis #DataQuality #DataObservability #DataEngineering #DataPlatform #CI/CD #DataGovernance #OpenLineage #ColumnLineage #DataContracts #ShiftLeftData #DataOps #AnalyticsEngineering #DataInfrastructure #BusinessAndTechnology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

A large European bank recently caught a data quality issue that would have misstated its risk-weighted assets by 8%. They didn't catch it with a rule-based check—they caught it with a machine learning model trained on historical data distributions. In this episode, Lucas and Luna explore how data teams are building ML models to monitor data quality across thousands of tables, detecting anomalies in freshness, distribution, and integrity that static thresholds miss. They discuss the mechanics of training on historical data, the challenge of false positives, integration with existing pipelines, and the shift in team skills required. The conversation covers hybrid approaches combining rule-based checks with ML, cost considerations, and the cultural change of trusting probabilistic detections. Lucas shares a practical starting point: pick your most critical datasets, use open-source libraries, and iterate. The episode closes with a look at the future of data quality monitoring becoming fully probabilistic.

#DataBusiness #MachineLearning #DataQuality #DataObservability #AnomalyDetection #DataScience #DataEngineering #BusinessTechnology #FexingoBusiness #BusinessPodcast #DataMonitoring #MLOps #DataGovernance #DataDrift #DataReliability #DataOps #DataStrategy #DataInfrastructure

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna explore how forward-thinking data teams apply product management principles to measure adoption, retention, and engagement of their internal data products. They walk through a case study of a large retailer that reduced time-to-insight by 40% by tracking metrics like weekly active users, query volume, and user satisfaction scores. They also discuss the role of data product managers, how to kill features that aren't serving users, and how to set service-level objectives for performance. The episode challenges the notion that 'if you build it, they will come' and offers practical frameworks for data teams to treat their outputs as products worth iterating on. After the break, they share a parallel example from a financial services firm that used net promoter score to improve their risk analytics product.

#DataProduct #ProductManagement #Analytics #DataTeams #DataAdoption #Metrics #DataProducts #RetailCaseStudy #BusinessPodcast #FexingoBusiness #DataStrategy #UserEngagement #DataDriven #ProductMarketFit #DataProductManager #FeatureAdoption #QueryLatency #CustomerInsights

Keep every episode free: buymeacoffee.com/fexingo

A major insurance company revealed that its analysts were spending nearly a third of their time simply finding and accessing datasets. In response, they built an internal data marketplace — a catalog-driven platform with data product pages, quality scores, and one-click access. The result: average time from request to delivery dropped from two weeks to two days, a 40 percent improvement in discovery efficiency. This episode unpacks the key components of a successful data marketplace, including governance integration, consumer ratings, and producer incentives. We explore how treating data as a product with its own storefront can transform data team dynamics and reduce friction across the organization.

#DataMarketplace #DataDiscovery #DataProduct #DataGovernance #DataCatalog #SelfServiceAnalytics #DataPlatform #Insurance #DataAccess #DataQuality #Analytics #Business #Technology #DataInfrastructure #DataTeam #DataDemocratization #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In episode 136, Lucas and Luna explore how data teams are adopting feature stores to solve the longstanding problem of feature engineering for machine learning. They break down the core idea: treating features as reusable, versioned, and governed assets — not one-off scripts. The discussion covers the open-source project Feast and the commercial platform Tecton, with a concrete example of how a fintech company cut feature deployment time from weeks to hours by implementing a feature store. The hosts also address common pitfalls like feature duplication and stale data, and explain why the feature store is becoming a standard part of the modern data stack alongside data warehouses and orchestrators.

#FeatureStore #MachineLearning #DataEngineering #MLOps #DataInfrastructure #Tecton #Feast #FeatureEngineering #DataGovernance #DataPipeline #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #Data #AI #DataTeam #DataStack #Metadata

Keep every episode free: buymeacoffee.com/fexingo

Episode 135 explores the rise of data contracts — formal agreements between data producers and consumers on schema, semantics, and SLAs. Lucas and Luna discuss how companies are reducing data incidents by 40% through contract enforcement, the cultural shift required, and why data contracts may become as standard as code versioning. They share a concrete example: a marketing team using contracts to guarantee a weekly churn score feed. The episode also touches on comparing data contracts to API contracts and the role of data platform teams. Plus, a brief listener-support appeal woven into the conversation.

#DataContracts #DataGovernance #DataQuality #DataEngineering #DataProducts #DataSLA #TrustInData #SchemaEnforcement #DataPlatform #DataIncidents #BusinessAndTechnology #DataManagement #DataObservability #DataLineage #DataCulture #FexingoBusiness #BusinessPodcast #Fexingo

Keep every episode free: buymeacoffee.com/fexingo

Most data catalogs are just glorified spreadsheets. But a growing number of teams are flipping the script with active metadata—where the catalog profiles data, detects changes, and enforces governance automatically. In this episode, Lucas shares concrete numbers from a fintech company that cut data discovery time from two days to twenty minutes, and how a midsize retailer reduced data quality incidents by 40 percent in three months. The hosts explore the cultural shift required, the role of automation, and why active metadata platforms are becoming the control plane for modern data operations. If you're a data leader looking to streamline discovery and governance, this episode offers a practical roadmap.

#DataCatalog #ActiveMetadata #DataDiscovery #DataGovernance #DataQuality #Automation #DataObservability #DataEngineering #DataCulture #MetadataManagement #Fintech #Retail #OpenSource #DataSLOs #PII #DataTeams #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Data observability has evolved beyond uptime monitoring—it's now a cost-cutting lever. In this episode, Lucas and Luna examine how a mid-size FinTech trimmed its Snowflake spend by 30 percent—over $650,000 annually—by applying query-level observability. They discuss the difference between traditional monitoring and cost-aware observability, the top five query patterns that drive overruns, and how teams can identify waste without disrupting production. The episode covers tools like Monte Carlo, Satori, and custom dashboards, and explains why data leaders are shifting from reactive alerting to proactive cost governance. Listeners will learn one concrete approach: using observability to attribute cost per query and per team, then setting budgets based on that data. A pragmatic take on turning observability from a monitoring expense into a savings engine.

#DataObservability #CloudCostOptimization #Snowflake #DataEngineering #FinTech #QueryOptimization #DataInfrastructure #CostAttribution #MonteCarlo #Satori #DataGovernance #CloudCosts #Analytics #DataTeams #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #DataProductivity

Keep every episode free: buymeacoffee.com/fexingo