I enjoy building software that scales from customer-facing applications and distributed systems to cloud infrastructure and AI-powered solutions.
Software Engineer with 3+ years of industry experience across full-stack development, backend engineering, and AI/ML. I've contributed to enterprise platforms serving thousands of users, optimized high-performance APIs, and built cloud-native and distributed systems using C#, .NET, Python, Angular, AWS, and modern AI frameworks. I integrate cutting-edge AI tools like Claude Code, OpenAI Codex, and GitHub Copilot into my development workflow to accelerate building and problem-solving
Led development of an end-to-end condo booking platform supporting 5,000+ users, reducing manual reservation processing by 40%. Designed and optimized SQL schemas, stored procedures, and secure RESTful APIs, integrated payment gateway services, and reduced query latency by 35%. Established unit and integration testing practices across the .NET and Angular codebase while leveraging AI-assisted development tools to improve code quality and delivery efficiency.
Built and deployed production AI systems for large-scale product intelligence and retrieval, including a CNN/ViT image classification pipeline achieving 99.47% accuracy across 42K+ SKUs and 8,500 images daily. Designed a multimodal RAG platform using LangChain, LLaMA 2, and GPT-4 Vision serving 1,200+ users with 1.4-second average response times, and scaled a Qwen3-VL vision-language pipeline processing over 320K images monthly.
Developed McDonald's Ready on Arrival platform using C# and .NET, enabling geofence-triggered order preparation that reduced customer pickup wait times by over 50%. Improved REST API performance by 30% through large-scale load testing and optimization, validating system behavior under 10K+ concurrent users. Investigated and resolved P1/P2 production incidents through root cause analysis and rapid deployments, improving reliability of customer-facing ordering systems.
Developed a machine learning solution for flight-train route prediction using Google Maps API and Random Forest models, achieving 92% accuracy and reducing manual forecasting effort by 70%. Built a scalable React application with Selenium-based automated testing, delivering real-time predictions with 99.9% uptime.
Developed a rumor detection system using Twitter API data, leveraging a Recursive Neural Network to classify true and false information with 83% accuracy. Enhanced model performance through a Branch LSTM architecture that improved contextual understanding of conversation threads and rumor propagation patterns.
Clinical decision support prototype for dermatology using a custom CNN with OpenCV. Classifies acne, eczema, and melanoma risk across 12K+ training images at 92.8% accuracy (MCC evaluated). Served 2,500+ users.
Engineered a multimodal RAG platform using Llama 3.2 and hybrid BM25+FAISS retrieval to answer questions across PDFs, DOCX, CSV, and image-based documents. Built a concurrent producer-consumer processing pipeline with mutex synchronization and task queues to scale document extraction, embedding, and retrieval workflows.
Architected a Function-as-a-Service platform using Redis Pub/Sub and ZeroMQ to enable asynchronous, event-driven task execution across distributed worker pools. Built fault-tolerant worker services with adaptive load balancing, backpressure control, heartbeat monitoring, and automatic failure recovery for resilient task processing.
Built an AI-powered DevOps assistant that automated the generation and validation of Terraform and Kubernetes configurations using OpenAI GPT. Containerized with Docker, deployed on AWS EC2, and integrated with GitHub Actions CI/CD and Grafana monitoring to streamline infrastructure provisioning and deployment workflows.
Built a scalable social media platform in Go featuring real-time feeds, post creation and editing, and threaded comments. Developed high-performance REST APIs on PostgreSQL, optimized for concurrent workloads, and improved security and responsiveness through JWT authentication, Redis caching, rate limiting, and email verification.