Skip to content
Muthu Bommiah

Initiative · USAA and Wells Fargo · 2020 – 2024

Controlled Data Movement

A configuration-driven, secure data movement pattern built at USAA to move 25B+ records a day, then carried to Wells Fargo as a bank-wide framework for analytics and AI.

25B+records per day at USAA
20xscalability in the next-generation platform
99.9%service-level objective

Problem

Analytics and machine learning need large volumes of sensitive data in the cloud, moved reliably, with every control a regulator would expect.

Approach

  • At USAA, architected a configuration-driven pipeline moving 25B+ records and 800,000 files a day into AWS and Snowflake.
  • Delivered a next generation with 20x scalability, 10x lower latency and a 0.01% exception rate at a 99.9% SLO.
  • At Wells Fargo, worked with data leadership to turn the blueprint into a bank-wide controlled data movement framework and archetype.
Data moves from a source zone to a target zone only through a highlighted governed path of classify, approve, protect and record lineage, steered by a policy engine. A dashed ungoverned copy path is blocked.Policy engineSourcezoneTargetzoneGOVERNED PATHClassifysensitivityApproveowner, purposeProtectencrypt or maskRecord lineagesource to targetungoverned copy: blocked
Data moves only through approved patterns. The governed path is the easy path, and every other route is closed.

Governance

  • Strict data isolation, granular access control and perimeter-based ingress and egress rules.
  • Column-level encryption and de-identification for sensitive data.
  • Automated audit logging, validation, rule-based reprocessing, lineage and catalog integration.

Outcome

  • A resilient data backbone for analytics and AI at two regulated institutions.
  • Observability, lineage and catalog integration built into the pattern from the start.

Lesson

Make the secure path configuration, not custom code, and teams will choose it.