Entering: Terminal

Where I code

My coding workspace. Projects spanning quantum transport simulators, distributed systems, machine learning pipelines, and algorithmic trading systems.

Quantum Transport

2 projects

scf‑negf — Self‑Consistent NEGF Quantum Transport Simulator

2024
•Developed a self-consistent NEGF + DFT solver with Anderson mixing convergence and PySCF integration for quantum transport calculations.
•Implemented iterative Poisson–Schrödinger coupling and matrix-based Hamiltonian construction using quantum chemical data to achieve self-consistent potential convergence.
•Analyzed current–voltage (I–V) characteristics and transmission spectra to evaluate quantum conductance in molecular junctions, validating computational results against analytical benchmarks.
•Designed a modular workflow adaptable for integration with machine learning–based surrogate models for accelerated quantum transport prediction.
RELATED PUBLICATIONS
→Quantum Transport from First Principles: A Self-Consistent DFT + NEGF Implementation in Python
PythonPySCFNEGFDFTQuantum TransportScientific Computing

pyRUQT — Recursive Quantum Transport Package

2023
•Implemented recursive Green's function algorithm with self-consistent corrections for efficient quantum transport calculations.
•Created a reusable package with comprehensive testing, documentation, and benchmarked performance optimizations.
•Developed modular architecture supporting multiple transport models and material systems.
RELATED PUBLICATIONS
→Quantum Transport from First Principles: A Self-Consistent DFT + NEGF Implementation in Python
PythonQuantum TransportGreen's FunctionRecursive Algorithm

Machine Learning

3 projects

LSTM Price Prediction Model

2024
•Built a deep learning model for financial time series forecasting using LSTM networks with attention mechanisms.
•Implemented comprehensive regularization techniques (dropout, L2 regularization, early stopping) and achieved improved generalization performance on out-of-sample data.
•Trained on multiple asset classes with hyperparameter optimization using grid search and cross-validation.
•Achieved robust performance metrics with focus on minimizing overfitting through careful feature engineering and model architecture design.
RELATED PUBLICATIONS
→Low-Rank Hankel SVD Spectral Learning for Regime Shift Detection in Financial Markets
PythonTensorFlowLSTMTime SeriesFinanceDeep Learning

Sentiment Analysis Transformer

2024
•Trained transformer-based models for financial news sentiment classification using BERT architecture fine-tuning.
•Achieved strong performance on sentiment prediction tasks with focus on financial domain adaptation.
•Implemented custom tokenization and preprocessing pipeline optimized for financial text data including earnings reports, news articles, and social media.
•Developed multi-class classification system with confidence scoring and attention visualization for interpretability.
RELATED PUBLICATIONS
→Low-Rank Hankel SVD Spectral Learning for Regime Shift Detection in Financial Markets
PythonPyTorchTransformersNLPFinanceBERT

Customer Behavior Analysis Platform

2024
•Developed a machine learning model to predict customer lifetime value, achieving 85% accuracy.
•Implemented a recommendation system with Cassandra backend for increased user retention.
•Utilized NLP techniques to analyze customer feedback, enhancing product insights.
•Built end-to-end ML pipeline with feature engineering, model training, and deployment workflows.
•Designed scalable data processing architecture for handling large-scale customer interaction datasets.
Pythonscikit-learnTensorFlowKerasCassandraNLPMachine LearningRecommendation Systems

Finance & Trading

1 project

Algorithmic Trading System

2024
•Developed a quantitative trading platform with comprehensive backtesting framework, risk management modules, and real-time data integration.
•Implemented multiple strategy classes including mean reversion, momentum, and statistical arbitrage with configurable risk parameters.
•Built portfolio optimization module with Sharpe ratio maximization and maximum drawdown constraints.
•Integrated real-time market data feeds with position sizing algorithms and automated trade execution with slippage modeling.
•Designed comprehensive risk management system with stop-loss, position limits, and portfolio-level risk controls.
RELATED PUBLICATIONS
→Low-Rank Hankel SVD Spectral Learning for Regime Shift Detection in Financial Markets
PythonQuantitative FinanceBacktestingRisk ManagementTradingPortfolio Optimization

Distributed Systems

1 project

Distributed Cache for Machine Learning Pipelines

2023
•Enhanced distributed cache system to expedite machine learning model access, achieving sub-5ms response time.
•Implemented custom monitoring for predictive model performance metrics using Prometheus and Grafana.
•Handled 10,000+ requests per second, simulating real-time data analytics and predictive scoring workloads.
•Designed write-through architecture to Google Cloud Storage for persistent caching and fault tolerance.
•Built horizontal scaling capabilities with Redis-based distributed key management and automatic failover mechanisms.
PythonGoogle Cloud ConsoleGCPRedisPrometheusGrafanaDistributed SystemsHigh Performance

AR/VR & Mobile

2 projects

SeaSOS - AR Evacuation Guidance

2024
3rd Place — NJIT HackTheLoc 2024
•Engineered an iOS app integrating AI and AR for indoor navigation, improving evacuation on cruise ships.
•Implemented Unity Navigation AI to calculate optimal evacuation routes, reducing time to exit by 20%.
•Employed RoomPlan API for real-time 3D mapping with navigation markers, achieving 99.9% accuracy.
•Built real-time pathfinding algorithms with dynamic obstacle avoidance for emergency scenarios.
•Designed intuitive AR interface with visual navigation cues and safety instructions.
UnitySwiftARKitRedisVuforia ARiOSARNavigationComputer Vision

MediSync - AI-Enhanced VR Health System

2023
1st Place — UPMC Hack 2023
•Designed VR-based real-time collaboration environment, leveraging AI to enhance digital healthcare experiences.
•Incorporated computer vision and ML for automated medical report and imagery analysis, achieving 98% precision in user-result matching.
•Built immersive VR platform for remote medical consultations and collaborative diagnosis.
•Integrated TensorFlow models for medical image classification and report generation.
•Developed MongoDB-based data management system for secure patient information storage and retrieval.
Next.jsMongoDBTensorFlowUnity VRGCPComputer VisionVRHealthcareAI

Web Development

3 projects

BlocSoc - NFT Authentication Protocol

2022
2nd Place — IEEE ProfHacks 2022
•Developed an NFT authentication protocol for apps to issue credentials proving participant identity on the blockchain.
•Built a decentralized blogging application with NFT-based authentication to promote free speech and censorship resistance.
•Implemented governance system where verified users can participate in content moderation, voting, and platform development.
•Designed merit-based token distribution system rewarding anonymous contributors based on community upvotes and participation.
•Created transparent, append-only blockchain ledger ensuring immutable records and permissionless access to public resources.
BlockchainEthereumNFTWeb3SolidityDecentralizedGovernanceAuthentication

SkillSwap

2025
3rd Place — Temple University OwlHacks 2025
•Designed a token-based community skill-exchange platform with a lightweight on-chain/off-chain blockchain rewards+trust model.
•Led UI/UX and data logic integration for seamless user experience.
•Implemented blockchain-based trust system for skill verification and community rewards.
•Built scalable platform architecture supporting skill matching and peer-to-peer exchanges.
•Developed decentralized skills marketplace empowering Philadelphia communities with token-based economy.
BlockchainWeb DevelopmentToken SystemUI/UXCommunity PlatformNext.jsTypeScriptReact

LinkWise Analytics Platform

2023
•Developed a data-driven link management platform, handling over 40,000 entities and user interaction data.
•Optimized data analytics pipeline with Tableau data visualization, resulting in 8% conversion rate improvement.
•Built scalable backend architecture with Node.js for high-throughput link processing and analytics.
•Implemented MongoDB data models for efficient storage and retrieval of link metadata and user interactions.
•Designed React-based dashboard with real-time analytics and interactive data visualizations.
ReactMongoDBNode.jsTableauData AnalyticsFull StackWeb Development