Capital One

Senior Manager, Software Engineering

CARD TECH - CUSTOMER SERVICING & STRATEGY

I was responsible for a team of software associates that support the Customer Service organization. So you know when you call the customer support line, it says; “your call may be recorded for quality and training purposes”? Well at Capital One it absolutely was. Every call was recorded in AWS Connect. My team was responsible for allowing a customer service agent’s manager to perform assessments on their calls. This assessment checked for any regulatory errors or other problems with the call. The system, called Sage, would allow the manager to use a set of hierarchical tags to mark places in the call to provide training and feedback to the agent. This system originally used Elasticsearch as its repository, but we upgraded it to use AWS OpenSearch. For this migration we moved more than five million customer contact records with zero errors, zero lost records, and zero downtime. This change allowed for closer alignment to the corporate tech stack (AWS) and made it easier to control system permission through AWS’ access management (IAM).

The Sage services are written in Java/Spring Boot and run on AWS instances. I upgraded the services to run in AWS Far Gate while moving simpler services to run in Node.js in AWS Lambdas (serverless).

We rearchitected the application to use a BFF pattern that reduced redundant intermediate services (~40% reduction in running ec2 instances) resulting in a faster responses and lowering overhead for maintenance and reducing vulnerability footprint.

Leadership

I worked with the multiple lines of business at Capital One and product owners to gather requirements, set business priorities (roadmap), and lead the team’s development and implementation of solutions.

Provided monthly reporting to senior management on subjects like incident management (frequency, severity, and duration from Service Now), ongoing project metrics (hours of new development, maintenance, and security/compliance from Tableau), and system metrics (assessments created, AWS spend, etc…).

Bi-weekly Architect Board to review any upcoming changes to the larger organization and how they interact with systems currently in place. This may be changes to the systems in Sage or even larger ones that affect multiple lines of business.

Other contributions

Worked on the OKR team to rescue overall costs. Part of this was in association with several of the projects listed above, like the migration to Fae Gate/serverless and applying the BFF pattern to reduce the number of running instances.