Episode 13 of The Data Business Podcast: Lucas and Luna dive into data observability — the operational discipline that separates companies that catch pipeline failures before they reach production from those that discover a broken revenue report at the quarterly board meeting. They explore why traditional monitoring isn't enough, how the five pillars of observability (freshness, volume, distribution, schema, and lineage) form an early warning system, and walk through a real 2025 case where a fintech company avoided a $70 million mis-pricing event thanks to a distribution anomaly alert. The hosts also unpack the economics: observability tools cost roughly 5–10 percent of a data engineering budget, but the cost of one undetected outage can be ten times that. Listeners learn the concrete difference between 'is my pipeline running?' and 'is my data trustworthy?' — and why the second question is the one that keeps chief data officers up at night. If you work in data infrastructure, analytics, or information products, this episode gives you a framework for convincing your organization that observability is insurance, not overhead.

#DataObservability #DataEngineering #DataInfrastructure #DataQuality #DataReliability #Monitorama #MonteCarloData #DataPipelines #AnalyticsEngineering #DataTrust #DataOps #BusinessPodcast #TechnologyPodcast #FexingoBusiness #BusinessPodcast #DataDriven #DataGovernance #DataCulture

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore how feature stores — centralized repositories for machine learning features — have moved from infrastructure nicety to operational necessity. They dive into the concrete case of a mid-sized fintech that cut model development time by 60 percent after adopting Feast, and discuss why feature reuse and consistency matter more than most teams realize. The episode also touches on the quiet tension between data engineering and ML teams, and how the feature store acts as a shared contract between them. With specific metrics and a real team's timeline, this is a grounded look at an infrastructure layer that's quietly becoming as important as the data warehouse.

#FeatureStore #MachineLearning #DataEngineering #MLOps #Feast #FeatureEngineering #DataInfrastructure #Fintech #Business #Technology #DataScience #ModelDevelopment #DataPlatform #FeatureReuse #TrainingServingSkew #DataGovernance #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore the rise of data unions — collective organizations that pool personal data to negotiate with tech giants. They examine the Swash union, which has aggregated browsing data from over 100,000 members to sell insights to ad platforms, and discuss the regulatory implications under GDPR and the EU Data Act. The episode drills into the economics of data as labor, comparing union bargaining to individual consent fatigue, and asks whether data unions can shift the power balance in the attention economy. Specific numbers include a 60% revenue share model and a projected $500 million market for personal data marketplaces by 2027.

#DataUnions #PersonalData #DataEconomics #Swash #GDPR #EUDataAct #DataAsLabor #AttentionEconomy #DataBargaining #ConsentFatigue #TechRegulation #DataCollectives #Privacy #DataMarketplaces #BusinessAndTechnology #DataInfrastructure #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Episode 10 of The Data Business Podcast tackles the messy reality of data mesh adoption. Lucas and Luna unpack Zillow's failed 2021 migration to a mesh architecture, why domain ownership created new silos instead of breaking them, and the surprising metric that companies should track instead of 'data quality scores'. They also discuss how one insurance firm got it right by pairing domain data stewards with a lean central platform team. If you're an analytics engineer, data leader, or CTO considering a mesh rollout, this episode gives you the concrete pitfalls and one counterintuitive fix: hire fewer data engineers, not more.

#DataMesh #DataArchitecture #Zillow #DataEngineering #Analytics #DomainOwnership #DataGovernance #DataPlatform #Business #Technology #FexingoBusiness #BusinessPodcast #DataPodcast #DataStrategy #DataQuality #DataSilos #DataStewards #PlatformTeam

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna examine how investment banks and corporate acquirers are now placing explicit dollar values on data assets during M&A transactions. Using the 2024 acquisition of Tableau by Salesforce as a reference point, they discuss a 2025 benchmark showing that companies with audited data quality scores commanded 12 percent higher multiples in private transactions. They also explore the emergence of 'data rooms 2.0' — separate virtual data rooms purely for due diligence on a target's data infrastructure, data lineage, and model training sets. Luna challenges Lucas on whether this is genuinely new or just a relabeling of existing intangible asset valuation. The conversation closes with a question about whether data valuation standards might eventually become as routine as GAAP accounting for auditors.

#DataValuation #MergersAndAcquisitions #DataQuality #VentureCapital #PrivateEquity #DataDueDiligence #Salesforce #Tableau #DataAssets #IntangibleAssets #AccountingStandards #DataGovernance #Business #Technology #DataInfrastructure #FexingoBusiness #BusinessPodcast #TheDataBusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore why data lineage — the ability to trace a data point from its source through every transformation — has become a regulatory and operational priority in 2026. Using the EU's latest financial reporting framework (ESEF 2.0) as a concrete case, they unpack how banks and fintechs are investing in automated lineage tools. Lucas cites a 2025 study showing that firms with mature lineage practices reduced audit costs by 18 percent and cut data incident response times by half. Luna challenges whether lineage is a genuine risk-management upgrade or just checkbox compliance. The episode also touches on the hidden tension between granular lineage and data privacy, referencing GDPR's Article 30. They wrap with a look at how open-source tools like OpenLineage and Marquez are shaping the vendor landscape.

#DataLineage #ESEF #FinancialCompliance #RegulatoryReporting #OpenLineage #Marquez #DataGovernance #GDPR #AuditReduction #DataProvenance #Fintech #DataInfrastructure #Business #Technology #FexingoBusiness #BusinessPodcast #DataEngineering #ComplianceTech

Keep every episode free: buymeacoffee.com/fexingo

Episode 7 of The Data Business Podcast dives into the real-world costs of testing data pipelines. Lucas and Luna examine a case study from a mid-size fintech company that discovered its data-quality testing suite was consuming 40% of its engineering budget—more than any other function. They break down the numbers: how testing time grew as data volume doubled every 18 months, the trade-offs between unit tests and end-to-end checks, and why many teams unknowingly over-invest in low-value tests. The episode also explores a counter-intuitive solution: reducing test coverage in exchange for faster iteration and lower costs. Lucas and Luna debate whether the industry's obsession with data reliability has created a testing tax that stifles innovation. Along the way, they reference real-world patterns from companies like Snowflake, dbt Labs, and Airflow users. The conversation ends with a practical framework for auditing your testing spend—something every data leader can apply immediately.

#DataPipelines #DataTesting #DataQuality #DataEngineering #Fintech #TestingCosts #Snowflake #dbtLabs #Airflow #DataInfrastructure #EngineeringBudget #DataReliability #CostOptimization #UnitTests #EndToEndTesting #DataVolume #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna dive into the quiet but tectonic shift happening inside modern data lakes: the rise of open table formats like Apache Iceberg, Delta Lake, and Apache Hudi. Lucas explains how these formats solve the decades-old problem of making cloud object storage behave like a transactional database, enabling atomic commits, time travel queries, and cross-engine compatibility. He cites Netflix's role in developing and open-sourcing Iceberg, which now powers over an exabyte of data in production at companies like Apple and Airbnb. Luna brings up the cost implications — how these formats reduce vendor lock-in and let companies swap query engines without rewriting pipelines. They contrast Iceberg's design philosophy with Delta Lake's tighter Databricks integration, and discuss the real-world operational trade-offs. The episode closes with Lucas's take on why format wars matter more than most engineers realize, and how the data lakehouse architecture is becoming the default for new data platforms. No jargon without explanation, no vague promises — just a clear, specific look at an infrastructure layer that quietly reshapes how data teams work.

#OpenTableFormats #ApacheIceberg #DeltaLake #ApacheHudi #DataLakehouse #DataEngineering #DataInfrastructure #Netflix #Databricks #CloudData #DataArchitecture #BigData #DataStorage #DataPlatform #BusinessAndTechnology #DataBusinessPodcast #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In Episode 5 of The Data Business Podcast, Lucas and Luna examine the rising use of synthetic data in financial modeling and the hidden risks of model collapse. Starting with a real-world case from a mid-2025 algorithmic trading blow-up, they explore how synthetic datasets—artificially generated to mimic real market behavior—can silently erode model quality. Lucas explains the statistical phenomenon of 'polya urn drift' where synthetic data causes compounding errors, while Luna questions the economic incentives for vendors to sell cheap synthetic alternatives. They discuss the JPMorgan pilot that found synthetic data reduced backtest reliability by 17% versus real data, and the new 'data provenance' standard emerging at the SEC. The episode drills into one concrete insight: synthetic data is not a substitute but a supplement, and knowing the difference defines a good data team. No sponsor breaks, no filler. Just sharp analysis for professionals who live in the data pipeline.

#SyntheticData #FinancialModeling #ModelCollapse #DataQuality #AlgorithmicTrading #JPMorgan #SEC #DataProvenance #MachineLearning #Backtesting #PolyaUrn #DataInfrastructure #BusinessAndTechnology #TheDataBusinessPodcast #FexingoBusiness #Analytics #DataScience #RiskManagement

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore the rise of real-time data feeds in commodities and futures trading, focusing on the shift from delayed exchange data to low-latency market data from firms like CME Group and ICE. They discuss how a 2019 change in wheat futures pricing data gave algorithmic traders a 15-millisecond advantage, forcing regulators to rethink market fairness. The episode drills into the economics of data distribution: exchanges now earn more from data licensing than from trading fees, and the rise of co-location services has created a two-tier information environment. Lucas explains why the SEC's 2024 proposal on market data transparency stalled, and Luna questions whether retail traders are permanently disadvantaged. A concrete example: the 2023 spike in natural gas futures during the Texas freeze, where firms with faster data feeds executed trades before the official price tick hit public screens. The conversation ends on the broader implication — data speed is now the primary competitive edge in markets built on milliseconds.

#RealTimeData #FuturesMarket #MarketDataFeeds #CMEGroup #ICE #AlgorithmicTrading #LowLatency #DataLicensing #SECRegulation #MarketFairness #CommoditiesTrading #FinancialDataInfrastructure #DataEconomics #NaturalGasFutures #WheatFutures #ExchangeData #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo