Full-Stack Data Engineer

Mission: To automate service level agreement tracking and provide real-time visibility into platform performance.

Engineered a full-stack monitoring system with Python data pipelines ingesting from Dynatrace API, Node.js/Express backend, and interactive dashboard frontend. Integrated multiple data sources from BitBucket and reduced manual reporting processes by 50%.

Mission: To identify key drivers of data consumption and optimize engineering efforts.

Processed multi-gigabyte Verizon RDD datasets using Pandas, reducing data volume by 90% and refresh latency by 60%. Identified top 10,000 high-usage VINs and delivered insights that guided engineering optimization efforts through interactive Power BI dashboards.

Mission: To build a secure, scalable, and modern full-stack e-commerce experience with seamless payments and dynamic management.

Developed a full-stack e-commerce platform using Next.js, MongoDB, and Tailwind CSS, integrating Stripe for secure payments, dynamic product and user management, and a scalable, modular architecture for future expansion.

Mission: To bridge healthcare and language gaps for multilingual families through accessible, AI-powered tools.

AI platform designed to improve healthcare access for non-English-speaking families. Integrates a bilingual health chatbot, Cora, medical document translator, and symptom-to-care matcher. Built with Vite + React, Tailwind CSS, Express.js, and OpenAI GPT-4 Turbo, it extracts text from medical PDFs using Tesseract.js OCR. Actively expanding multilingual support beyond Spanish, refining context-aware interactions, and improving OCR output clarity.

Mission: To make staying informed effortless and engaging through daily, personalized news quizzes.

Cross-platform quiz app that delivers personalized, daily news quizzes using TypeScript, React Native, Next.js, Tailwind CSS, and Firebase. Built in a monorepo with in-progress integrations for Stripe, PostgreSQL database, and AI-powered question generation.

Mission: To simplify API exploration and education through a sleek, centralized dashboard experience.

Web app that showcases interesting data from multiple public APIs. Built with Next.js, TypeScript, Tailwind CSS, and powered by a FastAPI backend with JWT auth. Deployed on Vercel (frontend) and Render (backend), the app integrates CORS-secured API communication, protected login flow, and dynamic UI rendering with real-time refresh.

Mission: To make self-discovery accessible and engaging by weaving together AI, zodiac signs, personality frameworks, and birth traditions into digestible, meaningful insights.

AI-powered reflection platform built with TypeScript, Next.js, and OpenAI. Delivers personalized journaling, secure auth, and user analytics. The app brings together a blend of culturally-rooted traditions around birth and identity. Presented in a way that’s visually engaging, emotionally resonant, and easy to understand.

Mission: To streamline construction workflows through domain-specific, AI-powered communication tools that enhance efficiency and decision-making on the job site.

AI-powered construction chatbot using Javascript, HTML, CSS, Vue, Electron, and Ollama for multi-model processing (DeepSeek R1, Qwen2.5).

Mission: To personalize meal planning using AI to promote healthier eating habits tailored to users’ preferences, lifestyles, and dietary goals.

AI-powered meal planning app that delivers personalized meal plans tailored to user preferences

Mission: To uncover insights behind user engagement through data visualization and machine learning—empowering education platforms to reduce drop-offs and improve retention.

Developed a Streamlit dashboard with interactive visualizations and SQL query functionality to analyze mock Duolingo learner data, applying machine learning models to evaluate user drop-off and engagement metrics.

Mission: To advance human-computer interaction by enabling accurate emotion detection using deep learning and facial expression recognition.

Developed a convolutional deep neural network for facial expression classification using PyTorch, achieving 1st place on the Kaggle leaderboard with a 0.82244 accuracy by applying advanced optimization techniques.

Mission: To provide instant, AI-powered first aid support and critical health guidance—making life-saving information accessible when it matters most.

Developed a full-stack medical chatbot web application using Flask, OpenAI, and Spotify APIs, providing instant assistance for medical inquiries with essential resources like CPR guidelines, first aid information, and emergency hotlines.