Hello!

I'm Varun Vekaria

A Full-Stack Developer specializing in TypeScript, React, and Node.js, building scalable, end-to-end applications on AWS. I work across the entire stack—from designing APIs and databases to cloud infrastructure and CI/CD pipelines—and enjoy transforming complex problems into clean, type-safe solutions. Driven to take on new challenges and ship impactful, production-ready products.

Resume Videofolio

Connect with me

Skills

Experience

Full Stack Developer — STUDYPAQ, SAN FRANCISCO (Remote)

05/2023 — 08/2023

Designed server-side architecture and constructed effective REST APIs using Flask framework and Postman for comprehensive testing ensuring 99% endpoint reliability.
Developed an ML algorithm and integrated Google APIs to create feature-specific video recommendation system catering to niche enterprise needs, serving over 10,000 video recommendations monthly.
Built a scalable pipeline to scrape large web applications by parsing their sitemaps and utilized Large Language Models (LLMs) to automate custom MongoDB database creation, saving over 100 hours of manual data entry.

Tech stack

Languages/Frameworks: JavaScript, Flask, Next.js, ReactJS, Node.js, MongoDB
Tools: Postman, AWS, OpenAI, Beautiful Soup, Selenium, Docker
Other: Software Development Life Cycle, Git

Machine Learning Intern — Rutgers Institute for Health

05/2025 — 08/2025

New Brunswick, NJ

Engineered a transcript parser with NLP techniques to process longitudinal interviews & extract 5k+ linguistic features.
Implemented a data pipeline handling 10+ edge cases, ensuring consistent inputs and generating a large-scale dataset.
Merged 14-day survey data from REDCap and MindLAMP, integrating additional features into a single database.
Visualized behavior trends using Power BI, enabling researchers to identify 3 fatigue patterns in post-COVID subjects.

Tech stack

Skills: NLP, Machine Learning, Data Pipelines
Tools: REDCap, MindLAMP, Power BI, Python
Domain: Healthcare Data, Longitudinal Studies, Data Visualization

Research Intern — INDIAN INSTITUTE OF TECHNOLOGY (IIT), PATNA

04/2023 — 10/2023

Contributed to the research of developing AI models for detecting phisher transactions in cryptocurrency.
Implemented baseline paper algorithms like Node2Vec, Trans2vec and GCN to benchmark model performance.
Employed Graph Neural Networks (GNN) and classification algorithms on Ethereum data and tested different feature generation techniques reaching phisher node detection accuracy of 98%.

Tech stack

Languages/Frameworks: Python (Machine Learning), Deep Learning
Tools: TensorFlow, PyTorch, Jupyter Notebook, Google Colab
Other: GitHub, Research Paper Skills, Presentation Skills

Front-end & Design Developer — UNICODE

10/2021 — 10/2022

Mastered core principles of UI-UX design and ReactJS, contributing to the successful launch of three user-friendly applications for an NFT marketplace and fitness sector.
Designed and implemented a feedback mechanism for HackPrep, collecting insights from 100+ attendees to refine future coding events; analysis led to enhanced learning experiences and session relevance.

Tech stack

Languages/Frameworks: HTML, CSS, JavaScript, ReactJS, UI-UX Design
Tools: MaterialUI, Ant Design Library, Figma
Other: GitHub, Agile Methodologies, Wireframing, Prototyping, User Research

Projects

Reconnect

Next.jsAWSTypeScriptOpenAIWhisperInsightFaceComputer Vision

Next.js–TypeScript web app with Flask backend for dementia patients: family-face recognition via InsightFace, memory captions with GPT-4o, and Twilio voice reminders.

GitHub

OmniMind

DockerLinuxCI/CDPythonOpenAIFastAPIAWSStreamlitNeo4jPostgreSQL

Agentic AI bias-mitigation system with FastAPI and Streamlit; MemMachine + Neo4j/PostgreSQL for context storage; multi-LLM comparison (OpenAI, Sonet, Gemini).

GitHub

EduProctor

Next.jsAWSFlaskInsightFaceComputer VisionNumPyOpenCVSnowflake

Built using Flask backend, integrating InsightFace for face detection, recognition, and embedding extraction. A Next.js + TailwindCSS frontend providing an intuitive interface for uploading classroom videos. Snowflake serves as the data store for embeddings and student metadata, while OpenCV and NumPy handle frame-by-frame analysis. The app runs end-to-end locally, processing classroom recordings into real-time engagement insights.

GitHub

My Portfolio

Next.jsReactTailwind CSS

This site—a responsive portfolio built with Next.js, featuring a dark theme and gradient accents.

GitHub

Stock Market Prediction

ReactJSFlaskTensorFlow

ML-powered stock price forecasting app with a React frontend and Flask API, using TensorFlow for LSTM-based time-series prediction.

GitHub

Kidney Abnormality Detection through Federated Transfer Learning

FlwrDistributed ComputingGoogle Colab

Federated learning pipeline for detecting kidney abnormalities from medical imaging while preserving data privacy across sites.

GitHub

GamePlan: Data-Driven NBA Schedule Optimization

PythonpandasDashPlotlyRegressiongeopy

Analyzed 10 seasons of NBA schedules with Python and pandas; built schedule-congestion features, Plotly choropleth travel maps, and XGBoost models to quantify schedule-driven performance impact.

GitHub

Trauma Analysis on Social Media

PythonDeep LearningOpenAI API

Multi-label classification of trauma types in social media text using deep learning and OpenAI APIs.

GitHub

Data Analysis of Stock Market

PowerBIPythonExcelER Diagram

Exploratory analysis and dashboards for market data using Python and Power BI, with ER modeling for structured datasets.

GitHub

Bot Chase Neural Network

Neural NetworksKerasBot Chase Game

Neural network agent that learns to play a bot chase game using Keras and custom environment design.

GitHub

Bot: Grid Fire Escape

PythonPath FindingPygame

Grid-based pathfinding game in Python with Pygame; implements classic search algorithms for escape behavior.

GitHub

RAG Project - Code Summarizer

ChromaDBLangChainOpenAI

Retrieval-augmented code summarization using ChromaDB, LangChain, and OpenAI for semantic search and generation.

GitHub

AI Agent - Daily Aggregated Bitcoin News

OpenAIAWSBrave APISupabase

Automated pipeline that aggregates daily Bitcoin news from the web and stores structured summaries via Brave API and Supabase.

GitHub

Publications

Stock Market Prediction using LSTM

(Algorithmic Trading, Version Control – Git, Financial Engineering)

Read

Multi-label classification for identifying trauma types on social media

(Predictive Analytics, Data Visualization)

Read

Detecting Kidney Abnormalities in Decentralized Healthcare via Federated Transfer Learning

(Distributed Computing, Medical Imaging)

Read

Achievements

2nd Place — MemMachine AI Agents Cross-Coast Hackathon (NYC–SF)

October 2025

JPM

SDE College to Corporate Program

J.P. Morgan · Apprenticeship · Jan 2022 – Feb 2023 · Mumbai, India

Hands-on Python, Django, and Linux training in backend development.
Built and deployed a live backend project; presented to J.P. Morgan professionals and received feedback on code quality and implementation.