Michael Legemah

- Reduced manual operational effort and accelerated customer time-to-value by architecting and deploying agentic AI workflows using AWS Bedrock, SageMaker, Lambda, AppSync, DynamoDB, OpenSearch, S3, and event-driven architectures, enabling automation of complex, multi-step enterprise processes at scale. - Improved production reliability and release confidence for AI-driven systems by designing and integrating LLM-as-a-tester and LLM-as-a-judge evaluation pipelines (Anthropic, OpenAI, NovaPro) into CI/CD, systematically detecting regressions, validating model behavior, and enforcing quality gates prior to deployment. - Increased system scalability and frontend delivery velocity by building GraphQL APIs with AWS AppSync to decouple frontend and backend concerns, improving data consistency, reducing integration friction, and enabling parallel development across teams. - Enhanced customer experience and accessibility across AI-powered touchpoints by delivering a production-grade contact center web application using React, TypeScript, Tailwind, Jest, and Python, ensuring UI consistency, performance, and WCAG-aligned accessibility at enterprise scale. - Strengthened platform security and system reliability by implementing secure authentication and event-driven integrations using Amazon Cognito, API Gateway, webhooks, and CloudWatch, supporting compliant, observable, and resilient distributed systems.

Clients: AstraZeneca, Evinova - Improved platform scalability and team throughput by leading the migration from a monolithic frontend to a modular React and Next.js architecture using TypeScript, Redux Toolkit, and Tailwind, enabling parallel feature development and long-term maintainability for clinical-trial platforms. - Eliminated manual documentation workflows and reduced turnaround time by designing and deploying LLM-driven RAG pipelines using LangGraph and AWS Bedrock, automating consent-form language translation while maintaining auditability and accuracy in regulated healthcare environments. - Delivered secure, high-performance backend services under strict compliance requirements by building REST APIs with Node.js, Fastify, and DynamoDB, meeting HIPAA standards for reliability, data protection, and availability. - Raised engineering quality and reduced defect rates by mentoring engineers in test-driven development and implementing automated testing pipelines using Jest, Cypress, and CircleCI, improving test coverage, accessibility compliance (WCAG, ARIA), and long-term code health. - Improved cross-functional delivery and product alignment by partnering closely with product, design, platform, and QA teams, translating regulatory and business constraints into scalable technical solutions.

Client: CACI Projects: US Army Artificial Intelligence Development Environment (AIDE) for Ai2C program ● Improved the UI performance on the US Army Artificial Intelligence Development Environment platform using React, TypeScript, and Redux. Got buy-in from the backend team to switch API from REST to gRPC. ● Spearheaded the refactoring of redundant React components by implementing AirBnB React Styleguide best practices, reducing code duplication by 30% and improving maintainability across the codebase, resulting in faster development cycles and enhanced team productivity. ● Identified the lack of unit and end-to-end tests on the front-end part of the codebase. Implemented Jest and Cypress for testing, and reached 90% code coverage. ● Collaborated with cross-functional teams on UI/UX architectural designs using Figma and ensured the implementation aligned with Application Security and WCAG compliance.

Projects: US Space Force Hyperion Project, Colorado Dept of Education Data Pipeline Modernization ● Led frontend development on the US Space Force Hyperion web application using React, TypeScript, Node, Material UI, Express, Lua, Python, MongoDB, Docker, and Kubernetes. ● Improved unit testing using Jest, Cypress, and Playwright, achieving over 80% coverage and enhancing application reliability. ● Led and mentored a team of 4 backend engineers on the Hyperion project regarding the frontend codebase. Mentored 2 front end engineers on the Colorado Dept of Ed project. ● Developed a robust front-end tech stack utilizing TypeScript, React, and Tailwind, accelerating UI prototyping for the Colorado Dept of Ed Data Pipeline Modernization Project, resulting in a 30% reduction in development time and improved user experience for educators and administrators. ● Promoted from Software Engineer III to Software Engineer IV due to outstanding performance in building resilient, high-performance web applications.

Projects: NYC OMNY, MBTA CharlieCard, Northrop Grumman ● Led a team of 6 engineers to build, launch, and maintain the Boston MBTA, NY OMNY, and SF BART Tap & Go payments web application using React, Tailwind, and GraphQL, resulting in an 80% increase in annual revenue. ● Developed and standardized Test Automation suites, increasing code coverage by 20% through CI/CD pipelines using CircleCI and GCP. ● Spearheaded the development of a comprehensive Front End Pattern Library for Northrop Grumman, utilizing React and PHP to create reusable WordPress blocks, resulting in improved design consistency, streamlined development processes, and enhanced site performance.

Clients: The Shadow League, 360i, AKQA ● Developed and implemented a Front End Pattern Library for the Mayo Clinic website redesign using Storybook, React, and Vue, resulting in a 90% reduction in bounce rate and significantly enhancing user engagement and navigation. ● Optimized the Shadow League website's performance during high-traffic sports events and podcast streaming, ensuring seamless user experience by scaling AWS EC2 instances and implementing AWS ELB, resulting in a 35% reduction in downtime and a 20% improvement in load times during peak traffic. ● Integrated and optimized premium ad placements using Google Tag Manager, increasing clicks and impressions by 85%.

Client: Pixacore

Projects: Juvenile Diabetes Research Foundation, National Marine Corps Society