Hi! I’m a recent University of Wisconsin–Madison graduate with degrees in Computer Science and Data Science, and a certificate in Economic Analytics. I have a keen interest in exploring new topics in the realm of computer science, data science and machine learning. I am a passionate problem solver, experienced programmer with an interest in Machine Learning and Artificial Intelligence. Passion, a strong work ethic, and attention to details are skills that I use on a daily-basis, which have propelled me to succeed in my work experiences to reach greater future goals.
Ark Dutt
69 W Squire Dr.
Rochester, NY, 14623, USA
(608) 960-5020
arksdutt@gmail.com
Computer Science and Data Science (B.S)• May 2025
Coursework: Data Structures and Algorithm, Data Modelling, Computer Architecture, Linear Algebra, Discrete Mathematics, Algorithms, Microeconomics, Intermediate Microeconomic Theory, Data Science Programming, Parallel & Throughput program, Data and Algorithms: Ethics and Policy, Machine Organization and Programming
International Baccalaureate• June 2021
Coursework: Maths AA HL, Physics HL, Chemistry HL, French SL, English SL, Economics SL
Mountain House Media • Aug 2025 - Present
I developed full-stack features for a SaaS platform designed to help creative professionals in media production manage and track their ongoing projects, using Next.js, Node.js, Express, and Python; improved workflow efficiency by an estimated 30%. I led the integration of Generative AI capabilities—including image and video generation, background removal, and upscaling—by developing and optimizing Python-based microservices and APIs, enhancing user productivity by 15%. Moreover, I also collaborated with stakeholders to define technical requirements, prioritize feature roadmaps, and mitigate delivery risks, ensuring the platform’s development aligned with user needs and creative workflows.
Wisconsin Center for Education Research • Feb 2024 - Jun 2025
I contributed to the ongoing ONPAR research project with an Agile team of six, focusing on enhancing an educational technology platform using a robust tech stack including C#/.NET, SQL Server, and AngularJS. I developed an automated scoring report system for over 100 science lessons by integrating a new database and collaborated effectively with a diverse group of stakeholders, including developers, designers, and educational researchers, to accurately capture and implement project requirements.
PBS Wisconsin • Sep 2024 - Jan 2025
I led a team of five developers to build and deploy an AR application using React.js, Flask, MySQL, and MindAR, integrating over 300 historical portraits to create a seamless 3D experience that increased user engagement by 20%. To enhance scalability and optimize cross-device performance, I leveraged MindAR and Three.js, reducing load times by 15%, achieving 99.9% uptime, and supporting over 50 monthly active users.
Informatics Skunkworks • Sep 2024 - Jan 2025
I conducted machine learning research on renal cell carcinoma (RCC) using CT images, focusing on improving tumor classification accuracy. My work involved the development of convolutional neural network (CNN) models specifically designed to segment over 160 RCC tumors, which contributed to enhancing clinical decision-making processes by providing more accurate and reliable imaging insights. Additionally, I optimized the feature extraction process from CT scans, resulting in increased efficiency of the data pipeline and improved overall model performance. These advancements have the potential to streamline RCC diagnosis and treatment strategies.
HeadStarter AI • Jul 2024 - Sep 2024
In this fellowship, I built over five AI applications and APIs using NextJS, OpenAI, Pinecone, and StripeAPI with a 98% accuracy rate. I was also leading a team of more than four engineering fellows where we developed projects from design to deployment using MVC design patterns. Additionally, I was being coached by engineers from Amazon, Bloomberg, and Capital One on Agile methodologies, CI/CD, Git, and microservice patterns.
TrustedNation • May 2023 - Aug 2023
In this internship, I spearheaded the development of both the backend and frontend components of the corporate website, restructuring the old frontend to a newer version using industry-standard tools like Next.js. This overhaul enhanced user experience and maintainability while improving and optimizing the website's rendering performance by 25%.
Software Training for Students • Nov 2022 - Jan 2024
I led several workshops and individual sessions focused on educating students in Data Science and Web Development technologies such as Python, R, HTML, CSS, and JavaScript. My dedication to high-quality instruction earned me an instructor rating of 4.80/5. Additionally, I created and refined STS workshop manuals, which led to a 75% improvement in student mastery rates.
Adira Finance • Jun 2022 - Jun 2022
In this internship, I managed the database of new automobiles using MySQL and utilized React.js and Node.js to handle both the backend and frontend of the website.
Beehive Drones • Jul 2019 - Jul 2019
During this internship, I gained experience on using C, Arduino, and Python to program microcontrollers and microprocessors by applied them to enhancing the performance of Internet of Things (IoT). Additionally, I assembled and tested electronic components of drones, contributing to their performance testing and simulation.
Developed a machine learning model to detect brain tumors given an MRI scan using SVM in Python. Implemented advanced image processing techniques using OpenCV for preprocessing MRI images to identify tumors. Achieved a classification accuracy of 95%, contributing to ongoing applications of AI in the healthcare industry.
Machine Learning, Data Science
Devised a machine learning model to accurately predict fraudulent transactions using Python and various libraries. Implemented Logistic Regression and various sampling techniques to assess the accuracy of the model. Produced a scalable solution to save significant amounts in fraud transactions and enhance customer security.
Machine Learning, Data Science
Developed an application in Java that allows users to search through posts on a Reddit forum using keywords. Worked on the backend to retrieve groups of posts that include specific keywords in their title or body. Helps users find the nutritional value of each food and distinguish which is healthy for them.
Programming, Java
Created a Java program that finds the shortest distance between two desired locations across the campus. Implemented the algorithm to find the shortest distance and calculate the energy required to walk. Assists students on campus to find the shortest way to go from place to place, especially freshmen.
Programming, Java
Designed a website using React.js and openweathermap API, which tracks the real-time weather of any city in the world. Provides various parameters like humidity, wind velocity, sunrise/sunset time, and the ‘feel-like’ temperature. Targets users who want to get information related to the weather of the place they are in.
Web Development, API
Conducted data analysis in R investigating different trends of car accidents in the state of New York. Displayed the proportion of accidents caused by in-state vs. out-of-state drivers and accidents of each vehicle type. Addressed the issue of accidents in New York and what the trend of accidents would look like in the future.
Data Science, R
Developed an AI-powered website with Node.js/Express, integrating Gemini Pro 1.5 (LLM) and IBM Watson APIs, handling 1000 interactions/month. Enhanced learning accessibility by creating an LLM-based interactive tool, improving user engagement by 40% and increasing retention by 30%, fostering deeper learning outcomes.
Headstarter AI Hackathon Finalist | Node.js, Express, LLM, APIs
Built an AI coding agent that autonomously generates full-stack applications from natural language prompts, streamlining end-to-end development using Python, LangChain, LangGraph, and the open-source Groq API (OpenAI-compatible), achieving 90% task accuracy across 25+ generated projects. Developed a Streamlit web interface and CLI tool for dynamic, real-time code generation with secure file orchestration and Pydantic-based workflows, reducing build time to under 45 seconds per project and improving overall efficiency by 3×
Python, LangChain, LangGraph, Streamlit, Groq (OpenAI-compatible) API, Pydantic