Michael Victor

Learn to build a feature Store in Fabric to prevent model reproducibility nightmares and promote feature reuse. Covers point-in-time correctness, feature versioning, and reusable schemas. For data engineers and data scientists building production ML.

Your data might be lying to you. Learn to detect Simpson's Paradox, the Will Rogers Phenomenon, and other analytical traps where aggregation and confounding variables flip conclusions. Walk away with practical strategies to spot and avoid costly mistakes

Fabric CI/CD tool selection doesn't have to be daunting. This session compares three deployment approaches: native Git integration, fabric-cicd/REST API, and Terraform across common patterns to help you choose the right tool for your needs.

Naming is more than labeling—it shapes what things become. This session explores naming conventions for tables, columns, variables, and artifacts, focusing on readability and clarity.

Semantic joins enable deeper knowledge discovery, but traditional methods fail at scale. Explore how LLMs can be used to address this by replacing brittle string metrics and rules with adaptive semantic understanding, reducing manual effort in entity resolution and deduplication