• Converted several Python services from Pipenv to Poetry for faster, more consistent dependency management

• Defined a new, extensible configuration format to allow external developers to easily create new integrations entirely within the ActiveCampaign platform

• Built new components in React to provide a step-by-step integration configuration experience for customers

• Reduced duplication of IAM Role Policies generated by developer tooling, allowing a single IAM Role to access more SQS Queues

• Eliminated several N+1 query problems, improving the response time of multiple frequently used API endpoints

• Updated an I/O bound API endpoint to use async/await pattern, increasing the maximum number of requests per second, and reducing the maximum scale of the API server

• Replaced limit-offset pagination with cursor pagination, significantly improving the peak response time of an endpoint serving data from a table with tens of millions of rows

• Improved automated text extraction accuracy using Google Cloud Vision, Tesseract, and AWS Rekognition OCR solutions

• Designed and implemented a highly parallel event-driven document processing pipeline to decrease business costs

• Utilized various Amazon Web Services including Lambda, Step Functions, and SQS to modularize and independently scale microservices

• Applied Test-Driven Development (TDD) in the implementation of Regex based text extraction from vehicle loan documentation

• Provided stakeholders with a real-time read-only ElectronJS application to view document classification and text extraction accuracy

• Defined infrastructure and service CI/CD using Terraform and Jenkins