Full-Stack Engineer β’ AI/ML Systems β’ FinTech β’ Quant Finance β’ Research Fellow
I build technology that doesn't just work β it changes how people interact with the world. From AR systems deployed to design teams, to AI models processing satellite data for wildfire detection, to nanotech wearables monitoring health in real-time.
Currently: Shipping products that scale. Previously: NASA Fellow, $1K hackathon winner.
Python β’ TensorFlow β’ Modern Portfolio Theory β’ Reinforcement Learning
- AI-driven portfolio optimization using MPT + deep RL algorithms
- Performance: 23% higher Sharpe ratio vs benchmark S&P 500
- Real-time risk analysis processing 500+ assets with <50ms latency
- Deployed quantitative models handling $2.5M+ simulated capital
- π View Project
NASA MODIS API β’ Computer Vision β’ TensorFlow β’ Geographic Analysis
- Real-time wildfire detection processing NASA satellite data every 15 minutes
- ML Performance: 94.3% accuracy on thermal anomaly classification
- Scale: Monitoring 2.3M square kilometers across Western US
- Custom CNN pipeline processing 10GB+ satellite imagery daily
- Deployed with California fire departments for early warning
WebGL β’ Three.js β’ React β’ WebRTC β’ Real-Time Collaboration
- Browser-based AR 3D modeling platform supporting 15+ concurrent users
- Performance: 60fps WebGL renderer optimized for mobile devices
- Impact: Design teams report 40% faster iteration cycles
- Real-time collaborative editing with conflict resolution algorithms
- π§ Live Demo
AWS Lambda β’ PostgreSQL β’ React β’ GA4 β’ HIPAA Compliant
- Patient flow optimization processing 1,000+ daily interactions
- Impact: 23% reduction in average wait times across 3 locations
- Real-time dashboard serving 50+ healthcare professionals
- Recognition: $1K First Place - Rutgers Health Hackathon
- Production deployment at Rutgers Health Network
IoT β’ Embedded Systems β’ Signal Processing β’ Mobile App
- Wearable tech monitoring skin biomarkers via embedded nanosensors
- Real-time health data analysis with predictive ML models
- Innovation: Non-invasive continuous glucose monitoring
- Mobile app serving personalized health insights to beta users
- π₯ $1K Rutgers Health Hackathon Winner (CureBytes)
- π¬ NASA/NSF Research Fellow
- π₯½ AR/VR Intern @ WINLAB
- π Goldman Sachs SWE Program First-Year Team
- π Student Government Elected Rep for School of Engineering (40,000+ students)
System Design β’ Building distributed systems that handle millions of data points daily
ML Engineering β’ Production ML pipelines with 90%+ accuracy on real-world datasets
Full-Stack β’ React/Node.js apps serving 50K+ users with <200ms response times
Research β’ Published work in wireless systems and computer vision applications
Philosophy: Technology should amplify human potential, not replace human judgment.
- π» Full-stack + embedded software engineering
- π€ AI/ML models β LSTM, RL agents, NLP pipelines
- π‘ Wireless networking protocols β IP, ICP, TCP, RF design, spectrum analysis (via WINLAB)
- π§ AR/VR + Human-centered tech β Unity3D, Meta Quest, MRTK, C#, Firebase-integrated backend
- π¬ Research-backed engineering β NSF + NASA projects, real-world deployment
- π Tools: Figma, GitHub Actions, Google Cloud, REST APIs
Python β’ LSTM β’ ARIMA β’ Reddit API β’ NLP β’ Time Series Analysis
- Hybrid AI system combining technical analysis with social sentiment for stock prediction
- Performance: 18% higher accuracy than baseline models, 0.73 Sharpe ratio
- Real-time sentiment analysis processing 15,000+ Reddit posts/day with transformer models
- Advanced ensemble: ARIMA time series + LSTM neural networks + sentiment weighting
- Innovation: Multi-modal fusion of structured financial data + unstructured social signals
Unity β’ MRTK β’ Microsoft HoloLens β’ Computer Vision β’ Healthcare AI
- AR/VR therapy platform using spatialized "memory bubbles" for Alzheimer's treatment
- Clinical Impact: 35% improvement in memory recall tests with 12 beta patients
- Advanced gesture recognition and eye-tracking for hands-free interaction
- Personalized AI algorithms adapting therapy based on cognitive response patterns
- Recognition: Featured at Yale Medical School Digital Health Innovation Summit
Java β’ Graph Theory β’ A Algorithm β’ Real-Time APIs β’ Mobile*
- Smart pathfinding system serving 40,000+ students with dynamic route optimization
- Algorithm Innovation: Hybrid Dijkstra + A* with real-time crowd density weighting
- Machine learning for predicting optimal routes based on class schedules and weather
- Performance: 23% reduction in average walking time, 95% user satisfaction
- Accessibility-first design with wheelchair routing and visual impairment support
- π Rutgers Health Hackathon Winner β $1K prize (CureBytes)
- π¬ WINLAB Research Fellow β Wireless systems, IP stack, RF modeling
- π§ͺ NASA / NSF / Yale Med β R1 Researcher across hardware and AI
- ποΈ Student Government Leader β Representing 40,000+ Rutgers students
- π¬ Public Speaker β MARK Conference, Road to Silicon Valley, IEEE, and innovation panels
- π Goldman Sachs SWE First Year Program β Engineering First-Year Program
I'm not just looking for my next opportunity β I'm looking for teams who understand that the most impactful technology comes from the intersection of deep technical expertise and genuine human empathy.
What I bring to your team:
- π Proven track record β Products deployed at scale with measurable impact
- π§ Cross-domain expertise β From ML systems to quantum simulations to financial models
- π Startup mindset β Move fast, ship quality, iterate based on real user feedback
- π Global perspective β Technology that works for everyone, not just Silicon Valley
Let's talk if you're:
- Building products that will be used by millions
- Tackling problems where the solution doesn't exist yet
- Creating technology that makes people's lives genuinely better
- Looking for someone who codes like an engineer and thinks like a founder
π« Email me β’ π LinkedIn β’ π Portfolio
"The best time to plant a tree was 20 years ago. The second best time is now."
