Abd-ul Khader

I own the data and analytics estate of the DC Metropolitan Police Department end to end: platforms, pipelines, governance and BI, with the technical direction set alongside the CIO. Before that I was lead data architect for Caterpillar's dealer and enterprise domains, principal engineer and software architect at Verisk, where I set the lakehouse standards for catastrophe modelling and built clean-room data sharing under GDPR and HIPAA, and the ingestion architect for a ten-petabyte, fifty-thousand-node Hadoop estate at Verizon. Earlier stops include Mastercard and Comcast. AWS published my design for the Police Department's data processing system. I have built every rung of the ladder for real: canonical domain models over order, inventory and payment systems, event-driven CDC pipelines, medallion lakehouses on Delta and Unity Catalog, warehouses on Snowflake and Redshift, governed data mesh with Collibra and Palantir Ontology, and the row-, column- and tag-level access controls that let regulated data be used at all. Most recently that means putting LLMs and agents on top of governed data with MLOps and responsible-AI guardrails, not on top of raw tables. A session with me is for a decision that is expensive to reverse: the target-state architecture, the platform choice, the governance and access model a regulator or auditor will ask about, the migration you get one attempt at, or whether your data is actually ready for AI. You leave with the decision made, the reasoning written down, and the next two moves.