Software Engineer | Full Stack | High-Performance Systems | Scalable Architect
Full-stack engineer with experience building scalable, user-focused 0→1 products. Delivered platforms serving 1,000+ users, AI-powered tools, and high-performance dashboards using React, Node, MySQL, and Rust/WASM. Strong execution, systems thinking, and product mindset.
|
Production-grade C++ web crawler achieving 50-100 pages/sec with fault tolerance and policy compliance.
🏆 Highlights:50-100 pages/sec throughput • 10× CPU gains via SIMD • Robots.txt compliance • RocksDB persistence | View Code | Documentation
- Created an open-source, cross-platform webpage analysis engine with 1,000+ downloads across Rust, WASM, CLI, and C++ interfaces — delivering actionable insights on SEO, accessibility, and performance.
- Enabled tailored quality evaluation for diverse web domains by designing a modular scoring system supporting 8+ analysis profiles (e-commerce, news, portfolio, etc.) and customizable thresholds.
- Attained ultra-fast performance of less than 100ms per page and 180+ pages/sec in batch mode by developing parallelized DOM parsing, connection pooling, and achieving a 36–40× throughput gain over sequential execution.y.
Cross-platform Rust analyzer with 1,000+ downloads deploying to 4 targets from a single codebase.
🏆 Highlights:1,000+ downloads • 36-40× speedup • 180+ pages/sec batch throughput • WASM + C++ bindings | Live Demo | GitHub | npm | crates.io
- Built a high-performance web crawler capable of fetching 100+ pages/sec, powered by multi-threaded scheduling and domain-aware rate limiting across 256 partitions for large-scale indexing.
- Ensured uninterrupted crawling with minimal data loss by implementing a RocksDB-backed persistence and crash recovery system, enabling safe checkpointing and auto-resume.
- Reduced average fetch time to 0.5–2.0s and maintained a low memory footprint of less than 20MB by optimizing the network stack with HTTP/2 multiplexing, connection pooling, and SIMD parsing.
🔧 Other Projects (Click to Expand)
ML-Powered Predictive Maintenance System | View Code
- Developed ensemble ML models (Random Forest, SVM) achieving 95%+ accuracy for fault classification across 5 motor states.
- Engineered advanced feature extraction pipelines optimizing inference latency to <50ms for real-time embedded deployment.
- Simulated complex motor behavior and fault scenarios using MATLAB Simulink enabling comprehensive testing with 10,000+ samples.
- Built scalable data infrastructure with end-to-end pipelines for predictive maintenance applications.
Full-Stack Social Media Platform | Live Demo | View Code
- Architected full-stack social media platform with RESTful APIs and real-time chat using Socket.io, supporting 500+ concurrent WebSocket connections.
- Implemented secure authentication using bcrypt and JWT, reducing login issues by 15% through improved UX flows and error handling.
- Designed MongoDB schema with optimized indexing achieving <100ms query response times and supporting horizontal scalability for 1,000+ users.
- Optimized React rendering with useMemo and useCallback hooks, implementing virtual scrolling for feeds supporting 1,000+ posts.
Intelligent AI Browser Companion | View Code
- Developed Chrome extension with React and Manifest V3, integrating TensorFlow.js for in-browser AI-powered text features improving productivity by 30%.
- Reduced extension bundle size by 25% using Webpack 5 optimizations including tree shaking, code splitting, and dynamic imports for faster load times.
- Implemented context-aware UI with seamless DOM integration and responsive design, supporting both light and dark modes with smooth transitions.
AI-Powered Interactive Music System | View Code
- Developed an interactive musical instrument using microcontroller, LDRs, and motion sensors with real-time audio synthesis.
- Programmed in C++ and Supercollider to generate dynamic tones based on real-time hardware interactions.
- Designed for user-friendly operation and seamless synchronization with other instruments.
I'm always excited to discuss projects, innovative ideas, or potential collaborations. Let's build something amazing!
"The best way to predict the future is to invent it." - Alan Kay