Amit Kumar

For Walgreens, I led a team of 4 developers to develop and integrate a new customer identity and access management platform (CIAM, in-house), backed by Okta IdP. Jointly, with Okta architects, drafted the process flow to move User data to Okta (only necessary attributes), allowing the new microservices to work with Okta. I developed microservices to hydrate Kafka topics with custom user attributes and preferences for the downstream APIs to consume from, and maintained the user profile and preferences schemas in the schema registry for data governance and integrity.

I temporarily acted as a lead for the State Privacy Laws (SPL) implementation team to lead the development team in designing and developing the features for the SPL stories. This included meeting with other teams to discuss potential solutions and overcome any shortcomings. I proposed and implemented solutions for the SPL project, which helped in implementing the “Rights” functionalities without compromising the compliance status.

Modernized the trading platform for one of the big banks in Canada, where a new architecture was to be implemented, from tightly coupled services to a loosely coupled SOA, and a new platform, Rancher-based Kubernetes, saw an influx of trade, market data, reset rate, securities, GED, and risk APIs.

I handled the migration of APIs from the existing Vanilla Kubernetes to Rancher Kubernetes, deploying them with Helm and managing the entire process with GitOps. I also led the teams through onboarding to our new trading platform.

I worked with the UI and back-end teams to fix issues with the new UI and newly built microservices; some of these microservices were rewritten by me in a parallel effort to migrate them from DB2 to MS SQL. I modified EOD scripts to work with the HashiCorp Vault Agent using AppRole authentication to write secrets to RBAC-controlled files that only the EOD scripts could read, thereby passing the Enterprise Security Assessment.

I migrated Shopper's Drug Mart's Crown Jewel application and central data repository (CDR) to GCP. Earlier, the application was deployed as distributed services on-prem, deployed on a WebLogic server cluster, which was not scalable and had lower throughput. I rewrote those services in Spring-WS to enable faster startup, scalability, and more maintainability, and deployed them to GCP using Helm and GitLab runners. The DB was migrated to CloudSQL for PostgreSQL using Flyway. Services started achieving hundreds of TPS, which was earlier 30-40 TPS, all by leveraging Kubernetes HPA in conjunction with better-designed applications for the cloud.