ThejasNarasimha Murthy

At Amazon, I’ve worked across large scale frontend platforms and distributed backend systems, building everything from ML pipelines to real time data workflows that support millions of customers and thousands of partners. I built an end to end ML recommendation pipeline that serves around 80 to 100 million recommendations a day. It runs on EventBridge, Glue, Spark, Athena, and SageMaker, and the whole process is fully automated. I transformed a 100 percent manual, ticket heavy workflow that took 2 to 5 days into an AppConfig and EventBridge based system that updates in under 10 minutes across 20 plus locales. I integrated metadata for 100K plus IoT devices, built a Lambda ingest pipeline, and added new APIs and UI components. This moved the system from weekly updates to near real time freshness. I replaced 24 hour Salesforce batch syncs with a CDC and EventBridge pipeline that updates in about 5 minutes. I redesigned DynamoDB and OpenSearch flows to keep search results fresh and reliable. I merged a standalone portal with 10K plus devices into the Partner Solution Finder. I rebuilt detail APIs, added AppConfig driven featured device metadata, and launched a new Featured Devices experience for 1K prioritized devices. I improved DynamoDB and Elasticsearch performance by 30 to 40 percent and cut EC2 compute spend by around 30 percent through benchmarking. I also moved static assets to S3 and added code splitting, reducing load times by 40 percent. In addition to development, I actively contributed to on-call rotations, led incident triage, maintained runbooks, and partnered with cross-functional teams to resolve live issues efficiently. I also supported new engineer onboarding, facilitated lunch-and-learns, and fostered a culture of shared learning and operational excellence across the team.