Work Experience

Sept. 2025 - Present

• • Architected and deployed production-ready driver distraction monitoring system using advanced computer vision and deep learning techniques, implementing facial landmark detection with Mediapipe and OpenCV to analyze drowsiness patterns, eye-aspect ratios, and head pose estimation for real-time distraction detection with 94% accuracy
• • Engineered custom YOLOv8 transfer learning pipeline for multi-class distracted behavior classification (phone usage, eating, fatigue), fine-tuning on 15K+ annotated frames and optimizing model inference to achieve sub-50ms latency on edge devices using PyTorch framework and ONNX runtime acceleration
• • Built production-grade modular alert system with Streamlit UI for live feedback visualization, implementing multi-threaded processing architecture with concurrent frame capture and model inference pipelines to maintain consistent 20+ FPS performance while reducing system resource utilization by 35%
• • Integrated SQLite database for logging distraction events and generating analytics reports, enabling fleet managers to identify high-risk driving patterns and improve driver safety training programs through data-driven insights

March 2025 - June 2025

• • Developed and shipped 15+ production-ready interactive UI components using React Native, TypeScript, and Next.js with server-side rendering, implementing responsive design patterns and accessibility standards (WCAG 2.1) that achieved 25% improvement in user engagement metrics and 18% reduction in bounce rate across mobile platforms
• • Engineered sophisticated 3D visualization system for real-time in-app advertisement previews using Three.js, WebGL shaders, and optimized Blender assets, implementing dynamic camera controls and lighting systems that reduced user preview setup time by 40% and increased ad conversion rates by 22%
• • Architected comprehensive testing infrastructure using Jest, React Testing Library, and Cypress for E2E testing, writing 60+ unit tests, 25+ integration tests, and 12 end-to-end test suites that increased code coverage from 52% to 80% and reduced production bugs by 35%
• • Collaborated with cross-functional teams in Agile sprints to deliver features 15% ahead of schedule, participating in code reviews, daily standups, and sprint planning sessions while mentoring 2 junior developers on React best practices

May 2024 - Aug. 2024

• • Architected and deployed full-stack analytics platform featuring real-time dashboards with interactive data visualizations using React, Chart.js, and D3.js to display campaign performance metrics including donation trends, volunteer engagement, and community impact statistics, improving stakeholder decision-making efficiency by 30% and enabling data-driven resource allocation
• • Engineered intuitive campaign management system using React, Next.js 14 with App Router, Tailwind CSS, and PostgreSQL backend, implementing role-based access control (RBAC), automated email notifications via SendGrid, and drag-and-drop campaign scheduling that increased deployment efficiency by 45% and reduced administrative overhead by 60 hours per month
• • Developed RESTful API using Node.js and Express.js with JWT authentication, implementing rate limiting, input validation, and comprehensive error handling that processes 500+ daily requests with 99.8% uptime
• • Optimized database queries and implemented Redis caching layer, reducing average API response time from 850ms to 120ms (85% improvement) and improving overall application performance for 200+ daily active users