MeetupWhat Does Best in Class AI/ML Governance Look Like in Fin Services?Pushed backCharles Radclyffe disputes the view that regulation should be the only reason organizations introduce AI governance controls.21:20
Guardrails, governance and security in 2020
19 sessions
MeetupWhat Does Best in Class AI/ML Governance Look Like in Fin Services?
MeetupWhy Data Scientists Should Know Data EngineeringClaimSecurity and compliance issues should be considered from the beginning of a project, especially when data includes personal health information or personally identifying information.43:03
MeetupKubeflow vs MLflowClaimMLflow's model registry provides governance for approving and managing models as they move toward production.9:59
MeetupBuild vs Buy an ML PlatformPushed backDiego Oppenheimer argues that replacing an existing system is not automatically valuable if it is working and meeting required security, compliance, upgrade, and delivery needs.21:00
PodcastA Conversation Around Feature StoresClaimFeature stores emerged as machine-learning systems scaled in the number of models, users, and feature dependencies, creating needs for predictability, observability, and governance.7:18
MeetupMLOps #37 When You Say Data Scientist Do You Mean Data Engineer? Lessons Learned From Startup LifeClaimData collection, governance, integrity, and reliability should be handled as an organization-wide responsibility rather than assigned entirely to a newly hired data scientist.12:07
MeetupOperationalize Open Source Models with SAS Open Model ManagerPushed backIvan Nardini agrees with Demetrios Brinkmann that security is one of the hardest parts of building an open-source end-to-end machine-learning product.17:03
When Machine Learning Meets Privacy, Episode 1ClaimFabiana Clemente hosts the podcast series on data privacy for machine learning, powered by MLOps Community and sponsored by YData.0:00
What are regulations saying about data privacy?Pushed backDemetrios Brinkmann suggests that privacy regulations may be taken less seriously by smaller and disruptive companies, and Cat Coode agrees that this is often the case.8:08
MeetupUN Global PlatformClaimMark Craddock said the final United Nations platform was multi-cloud because no single cloud met the needs of users globally, partly for geopolitical and security reasons.6:56
Are Privacy-Enhancing Technologies a Myth?Pushed backThe Facebook project experienced a trade-off in which adding more randomness improved privacy but reduced data quality and utility.21:09
The intersection between DataOps and privacyClaimDataOps immutability can conflict directly with privacy requirements to forget or correct personal information.20:53
ML and Encryption: It's All About Secure InsightsClaimSecure multiparty computation can split a secret into random shares so that no individual participant can reconstruct it alone.5:21
Privacy-preserving ML with Differential PrivacyPushed backFabiana Clemente frames differential privacy mainly as a trade-off between utility and privacy, while Christos Dimitrakakis adds that privacy-induced randomness can sometimes improve generalization.12:28
PodcastDeep in the Heart of DataClaimCarl Steinbach says security becomes more tractable when users access tables, records, and views through an intermediate record layer instead of accessing the storage file system directly.43:36
MeetupHow To Move From Barely Doing BI to Doing AIClaimJoe Reis says companies often struggle with machine learning because they lack consistent data definitions, reliable reporting, data quality, and data governance.8:21
