• Introduced automated ML-powered title recommendations for Fire TV homepage, increasing platform’s streams annually by 26.72 million and viewing duration by 1.002 minutes per customer

• Architected and deployed a solution to fix a prod error showing duplicate movie titles by enabling communication between different services across the Prime Video stack through tokenization; introduced enhanced testing and monitoring to proactively prevent similar errors

• Led cross team initiative to improve Fire TV Movie and TV page UI, enhancing customer experience for over 200,000 customers per day

• Leveraged LLM models to improve Prime Video’s search capabilities on non-title searches and searches greater than 5 words in length

• Bolstered the operational efficiency of large scale Java applications (handling > 50,000 transactions per second) by introducing caching, throttling, and timeout and retry strategies

• Analyzed ticket data to identify common operational pain points and developed an action plan to address the shortcomings to improve overall operational state; internal ticket efficiency metric rose from 56% to 96%

• Drove development of junior engineers by hosting weekly tech talks and knowledge sharing on various topics such as AWS, Kotlin, and Amazon internal architecture/microservices

● Developed and improved ETL pipelines responsible for Capital One Credit Card Fraud data by using PySpark for application code, and Terraform for infrastructure deployment on AWS

● Leveraged AWS EC2 and Python to perform data validation on files that teams data pipelines produced, ensuring our production process ran correctly

● Made modifications and improvements to python module capable of scanning and masking millions of records for sensitive customer information

● Contributed features and enhancements to microservices based SpringBoot Java application hosted on AWS ECS. App provided enterprise wide solution for performance testing applications

● Engineered a solution to provide auto scaling down on idle AWS infrastructure using CloudWatch alarms and Lambda functions, saving the company money on AWS costs

● Improved application performance by modifying API and database query. Application that previously took greater than 10 seconds to load results now returns information within 100 ms

● Aggregated massive Elasticsearch index into smaller indices, reducing storage costs. Utilized Kibana and Javascript to generate visualizations and graph information

● Drove cross platform integration of two separate services, then ran weekly training meetings to instruct users, enabling over 200 teams within Capital One to onboard to our product

• Developed a webpage that allows users to verify the installation of specific Qualtrics features on a target URL through the use of Javascript postMessage and event handler functions

• Implemented AngularJS directives to detect updates to specific values on a webpage made by the user, and to verify if user input was valid by matching it against regular expressions

• Utilized Jasmine to write unit tests and end to end tests that verified my code functioned correctly

• Interviewed job candidates and evaluated their potential contributions to the team

• Maintained an effective and efficient communication method for over 80 employees through emails, social media and announcements, ensuring full transparency within the department