Desean Abraham

Designed and deployed multi-container ECS Fargate services, including a dual-container (Mirth Connect + web admin) and a five-container service (Artemis, web, web admin, two integration services), supporting patient intake for U.S. Army, Navy, and other U.S. military organizations.

Automated infrastructure provisioning and deployment for these containerized services using Terraform, significantly increasing deployment efficiency and consistency.

Configured VPN and network access controls to establish secure connectivity for the Buffalo office cloud environment, ensuring robust data security and compliance.

Led an end-to-end rebuild of a photo lineup feature used by Louisiana police officers, redesigning the SQL database schema and refactoring photo lineup API endpoints to reduce network latency.

Built React and Ant Design modals on the photo lineup view, enabling officers to efficiently view and edit suspect metadata via corresponding CRUD API endpoints.

Improved the suspect search experience by implementing pagination and loading states, enhancing usability and reducing query times.

Implemented multi-file upload via React hooks and the FormData API, with backing endpoints to securely store suspect photos in PostgreSQL.

Built a token-bucket rate limiter for the document management API, protecting the service under sustained load and enabling tiered API subscriptions.

Designed and deployed a scalable document reporting pipeline (AWS SQS, Node.js) with automated delivery and S3-backed storage for receipts, invoices, and mortgage documents processed by an AI model.

Architected an end-to-end queuing system (AdonisJS/React, RabbitMQ, AWS SQS + DLQ) that routed low-confidence AI-processed documents to a human review queue, including a React interface for queue statistics and configurable routing criteria.

Developed a React component enabling users to generate and download CSV reports across various document types, streamlining reporting processes through generalized and specific handling algorithms.

Built an AWS Lambda function to support SFTP document uploads, feeding them into the queue for AI-engine (JET) processing, alongside existing API and web-app upload paths.

Served on the engineering interview committee, conducting technical interviews that contributed to several full-time hires.

Built a shared calendar in React for patients and physicians, enabling real-time tracking of availability and upcoming appointments.

Developed a Reports component for physicians to aggregate and view patient PHI, enhancing data accessibility and compliance.

Created API endpoints and a reusable React component for printing and managing ZPL labels, streamlining test tube sample tracking.

Designed and built a customer subscriptions page in React for the Laboratory Information System Management product, clearly presenting available pricing models.

Delivered 'Contribution Breakdown' feature (AngularJS, .NET/C#, Entity Framework Core), giving 6,000+ employees detailed visibility into employer benefit fund allocation.

Built an allocation-bucket feature, informing employees of unused benefit funds and guiding them to maximize tax savings for the benefit year.

Configured real-time alerting (PagerDuty, Terraform, Azure Event Hub) to notify on-call engineers of HTTP 4xx/5xx errors, improving system observability and incident response.

Automated a manual SQL insertion process by building a Node.js script that converted product-manager-supplied translation spreadsheets into ready-to-run SQL, eliminating hours of repetitive manual work each release cycle.

Built a real-time car auction display in React and Python/Django with advanced search functionality, improving user experience.

Contributed frontend features and bug fixes to the internal admin site, replacing manual spreadsheet workflows and enhancing operational efficiency.

Worked with ElasticSearch to audit and track live car auctions, identifying and resolving issues that led to premature endings or critical response messages.

Led an internal hackathon win by building a sentiment analysis tool to surface and aggregate dissatisfied customer and competitor signals from Twitter/X in real time, enabling proactive customer support and automated replies.

Developed a hotel recommendation chatbot that utilized past flight history to personalize suggestions, improving user engagement.

Refactored flight search results using Test-Driven Development (TDD) to ensure accuracy and data freshness.