My journey through academics, achievements, and the passions that drive me forward.
A showcase of my work spanning software engineering, machine learning, and quantitative research.
Bridging the gap between athletes and personalized coaching
A comprehensive machine learning platform that predicts Formula 1 race outcomes using historical data, real-time statistics, and advanced ML algorithms with a modern web interface for interactive predictions.
Built a predictive modeling system to forecast cryptocurrency arbitrage opportunities using machine learning models.
A full stack application that makes use of websockets through multiple APIs to showcase the arbitrage opportunities in the cryptocurrency market for certain coins.
A project that explores algorithms for finding maximum matching in d-partite graphs, implemented in Python with comprehensive testing.
A collaborative project using R to build economic models for predicting inflation rates using TSLM and ARIMA models.
A custom programming language developed using OCaml with interpreter for basic programming language features.
Design and implementation of a 16-bit CPU using Logisim with support for arithmetic, branch, and jump instructions.
Exploring different LLM prompting strategies for answering questions about tabular data using various evaluation approaches.
Developed a chatbot using LLM API with retrieval-augmented generation and advanced prompt engineering techniques.
A comprehensive NLP toolkit for movie review sentiment analysis and complex word identification using multiple machine learning algorithms.
An alternative to Gradescope with enhanced features for educational institutions and improved user experience.
A comprehensive fitness tracker web application that enables users to plan workout routines and monitor their progress.
A user-friendly web application for efficient task and project management with drag-and-drop calendar interface.
Showing 16 of 16 projects
incoming software engineer working on embedded payroll
Databricks Lakeflow Connect architecture for data processing of real time and batch events. Purpose build pipeline for BI, Analytics and ML use cases
Built scalable analytics pipelines using Databricks Lakeflow Connect, transforming raw data into business-ready insights for product and growth teams
Built serverless data pipeline using AWS Lambda, S3, and FastAPI that automates complex file merging workflows, reducing processing time by over 95% (1 hours to 3 minutes) and eliminating manual errors. Developed full-stack React TypeScript application with real-time polling and drag-and-drop interface, processing files up to 10GB while maintaining responsive UX and 99% success rate. Engineered ETL data transformations with Pandas to merge multi-format datasets (CSV/Excel), implementing business logic and aggregation algorithms that save 15+ hours weekly of manual work.
Developed a predictive ML pipeline using scikit-learn and statistical methods to train predictive models (Random Forest, SVM, Gradient Boosting) to forecast optimal arbitrage strategies with 86.04% accuracy and simulate $897K+ trading profit across 30K+ trades. Engineered a time-series dataset of 28M+ BTC/USD ticks using Pandas and NumPy for feature extraction.
Built a full-stack real-time crypto analytics platform using Flask, WebSocket and CoinAPI, streaming BTC/USD data from several exchanges with <200ms latency to detect and visualize arbitrage opportunities. Designed and deployed an end-to-end ETL pipeline to serve real-time price data via REST API and WebSocket.
Led the creation of 1000+ tests to validate d-partite graph matching algorithms, improving reliability and identifying edge case errors. Applied advanced algorithm design techniques to improve performance and reliability of maximum matching in d-partite graphs, achieving 99% improvement in theoretical performance.
Engineered comprehensive ETL pipelines in Python and R using Pandas and NumPy to process and transform 1.5M+ shipping records for business intelligence reporting. Designed and implemented data warehousing solutions with Firebase and Firestore, creating scalable analytics infrastructure that reduced data processing time by 40%. Developed interactive BI dashboards and data visualization tools using React and Google Maps API to provide actionable insights for logistics operations.
Developed and optimized Python scripts to automate data collection and processing workflows using Object-Oriented Programming (OOP) principles, resulting in a 30% reduction in manual processing time. Collaborated with Senior programmers to build Python-based executables, applying OOP design patterns.
A live banner of languages synced from Convex.
10 languages
Auto-scroll pauses on hover for readability.
The technologies I work with daily, from frontend frameworks to cloud infrastructure.
24 technologies · hover to pause
Infrastructure & Certifications
Let's discuss opportunities, collaborations, or just have a chat about technology.
Currently open to full-time positions and interesting collaborations.
Built with passion in 2026