I'm developing an AI construction platform that revolutionizes project management through predictive analytics.
From NASA satellite systems to construction AI startups, I build scalable solutions that solve real-world problems. Currently pursuing Computer Science at UW-Madison with hands-on experience in full-stack development and machine learning.
Here's a quick snapshot of who I am and what I do
I'm a Computer Science student at the University of Wisconsin - Madison with real-world experience building systems that matter. From processing satellite data at NASA to creating AI-powered solutions for construction, I love tackling complex technical challenges.
My expertise spans full-stack development, machine learning, and system architecture. I've worked with everything from AWS cloud infrastructure to neural networks, always focusing on creating scalable, efficient solutions.
Currently, I'm building an AI-powered construction startup while finishing my degree. I believe technology should solve real problems and improve people's lives - that's what drives every project I work on.
Expected Graduation: May 2026
Cumulative GPA: 3.72
Relevant Coursework: Object-Oriented Programming, Data Structures,
Algorithms, AI, Operating Systems, Software Engineering, System Architecture for
Quantum Computers, Machine Organization
Architected real-time satellite tracking system on AWS EC2, processing daily thermal infrared data and reducing processing latency by 75%. Built neural network for cloud phase detection achieving 95%+ accuracy. Developed full-stack web application improving mission operations efficiency by 40%.
Developed production-grade practicum management system using Node.js, Express.js, MySQL serving 100+ concurrent users with 99.9% uptime. Engineered Python-based priority matching algorithm reducing administrative tasks by 60% with 100% successful placement rates.
Building and maintaining custom Shopify e-commerce solutions using JavaScript, JSON, HTML/CSS and Liquid. Developing interactive product customizers and creating responsive designs across desktop and mobile. Optimizing website performance and translating business requirements into technical implementations.
Assist 100+ students in understanding core AI/ML concepts, including search algorithms, reasoning, and machine learning. Guide students through programming assignments in Python, focusing on AI applications such as natural language processing, game theory, and neural networks. Collaborate with instructors and TAs to identify common student challenges and improve course materials.
Conducted research on generative models for predicting materials' structure and properties. Developed scripts to visualize structures and optimized model performance using subsets of data and Langevin dynamics steps. Collaborated on analyzing performance metrics including RMSE and Match Rate.
Oversaw student staff work, conducted facility inspections, and performed maintenance tasks. Trained staff on equipment installation and maintenance while ensuring compliance with safety protocols. Led priority repairs during summer conferences and prepared buildings for occupancy.
Designed data-driven inventory management solutions, reducing stockouts by 15% and enhancing data processing speed by 30%. Developed an AI-driven predictive model for inventory forecasting, reducing overstock by 20%. Collaborated in Agile teams to create intuitive user interfaces using React.
Spearheaded data collection efforts across multiple branches using Python. Utilized advanced statistical tools, reducing data processing time by 30% and providing actionable insights. Applied Python and Pandas for in-depth data analysis and cleaning processes.
Technologies and tools I work with
Check out some of my recent work
AI-powered course assistant for 45,000+ UW-Madison students using Google Gemini AI, Python Flask, and Chrome Extension. Achieved 99% parsing accuracy with intelligent query routing and zero failed requests.
Real-time satellite tracking system for NASA PREFIRE mission on AWS EC2. Built neural network for cloud phase detection achieving 95%+ accuracy. Full-stack application improving mission operations efficiency by 40%.
Custom LSTM model in PyTorch to forecast receipt volumes using historical time series data, with an interactive dashboard featuring dynamic visualizations and parameter tuning.
Developed an interactive product customizer for Shopify allowing customers to design personalized pet collars with various patterns, colors, and text options.
Full-stack web application enabling real-time song requests at DJ events with responsive interfaces for both DJs and attendees using HTML, CSS (Tailwind), and JavaScript.
A text-based RPG game using Convex for game logic, ChatGPT API for dynamic prompts, and DALL-E for image generation.
An Android application aimed at managing food waste and facilitating donations, increasing user engagement in the initiative.
Led the Frontend development of a dynamic application to enhance dealership-used car inventory management with RBT algorithm.
Feel free to reach out for collaborations or just a friendly chat
krishangmittal2004@gmail.com
+1 6087077944
Madison, Wisconsin
Looking for feedback and potential collaborators!