Team: Query Infrastructure

• Lead development on next generation batch processing system, reducing batch processing costs by 50% while also decreasing processing times and on-call burden (reduced batch SLA from 3 days to 25 minutes, without increase in incidents)

• Optimized Dynamo operations from 100k+ units/s bursts to <3k units/s bursts during heavy load

Team: Query Infrastructure

• Built memory limit detection for user-centric queries and automated killing and splitting of problematic queries to ensure system stability for all tenants

• Created new fairness rules to prevent singular users within one database from bottlenecking all other computations within the database, reducing average throttle wait times before query computation by 30%

• Optimized the memory and CPU usage of various computations to reduce GC overhead and increase throughput

Team: Query Infrastructure

• Reduced customer chart latencies by up to 80% by analyzing load patterns, improving caching performance, optimizing query cost calculations to reflect real-world resource usage, and shuffling load around query clusters

• Rewrote real-time user-based event ingestion service to process 6x more load with 20% less machines

Team: Abuse Prevention

• Designed and developed full-stack application to automate work that used to take weeks to take 2 days

• Interfaced with several internal services to achieve project goals

• Utilized Angular framework to create responsive front-end

• Created REST API that used AWS API Gateway and Lambda, with Cognito and OpenID for authentication

• Deployed project via CI/CD, utilizing AWS CloudFormation to generate infrastructure

• Participated in agile development process

• Utilized knowledge in Angular, Lambda, API Gateway, S3, CloudFormation, IAM, Cognito, TypeScript, Python, and Git

– Successfully rewrote problematic legacy Fortran code to Python that had been attempted for years without success

– Wrote REST endpoints with Java and Spring to save, retrieve, and manipulate CSV data for MySQL

– Created React views and updated existing views to interact with REST endpoints

– Managed and maintained data storage infrastructure that held data from experiments

– Created continuous integration pipeline to automate project deployment to Amazon EC2

– Wrote script to automate installation and configuration of Torque for a single node on EC2

– Leveraged knowledge in Java, Python, Spring Boot, JavaScript, React, MySQL, REST, Git, Gradle, Docker, Linux, Bash, CircleCI, AWS, and Chrome Developer Tools