Full-Stack & High-Throughput Media Platform Engineer at Civiconnect (specialized workforce and technology accelerator)

Engineered a high-availability, full-stack media processing platform that sustained a peak workload of 1000+ daily active image classification requests with ~200ms average end-to-end response time. Developed a robust RESTful API using Node.js/JavaScript to orchestrate pipelines and route client traffic to TensorFlow models for real-time analytics. Built a responsive web dashboard with React and CSS, incorporating asynchronous file uploads and real-time telemetry visualizations for large user-submitted image payloads. Designed and implemented a relational MySQL schema for user metadata and high-frequency classification history, using indexing to reduce query execution time by 35%.

Full-Stack & High-Throughput Media Plat at Civiconnect (specialized workforce and technology accelerator)

Engineered a high-availability full-stack media processing platform supporting 1000+ daily active image classification requests with ~200ms average end-to-end response time. Built a RESTful API using Node.js/JavaScript to orchestrate pipelines and route client traffic to TensorFlow models for real-time data analytics. Developed an interactive web dashboard using React and CSS with asynchronous uploads and real-time telemetry visualizations for large image payloads. Designed a relational MySQL schema for user metadata and high-frequency classification history, using indexing to reduce query execution time by 35%.

Full-Stack & High-Throughput Media Platform Engineer at Civiconnect(specialized workforce and technology accelerator)

Engineered a high-availability full-stack media processing platform sustaining a peak of 1000+ daily active image classification requests with ~200ms average end-to-end response time. Developed a robust RESTful API using Node.js/JavaScript to orchestrate pipelines and route client traffic to TensorFlow models for real-time analytics. Built an interactive, responsive web dashboard using React and CSS with async uploads and real-time telemetry visualizations for large image payloads. Designed and implemented a MySQL relational schema for user metadata and high-frequency classification history, adding indexing to reduce query execution time by 35%.

Machine Learning & Full Stack Engineer at Toronto Metropolitan University

Architected a dual-engine Python Flask backend providing concurrent REST endpoints for real-time comparative inference. Implemented a high-performance Python client to stream high-throughput image data via REST APIs. Built a dynamic client dashboard using vanilla JavaScript and HTML to handle asynchronous image streaming and render visual feedback alongside model confidence scores. Developed an end-to-end computer vision platform using transfer learning with TensorFlow MobileNetV2 and HuggingFace ResNet-50, achieving 92% classification accuracy through model tuning and evaluation.

Cloud-Native AI-Powered Job Matching Platform at Ministry of Public and Business Service Delivery and Procurement

Architected and implemented a scalable, event-driven resume-to-job matching platform using React.js and asynchronous Node.js. Designed an asynchronous data pipeline with Amazon EventBridge to decouple job ingestion from AI analysis workflows. Implemented heavy-lift computation in stateless AWS Lambda functions to reduce API blockages and improve responsiveness by ~30%. Enabled secure, high-throughput resume uploads via API Gateway endpoints that generate AWS S3 pre-signed URLs, offloading large uploads directly to cloud storage.