Andres’ Portfolio
A bit about me, my work experience, extracurriculars, and personal projects is on this website. Click below to jump to a section or simply scroll through.
About Me
Interests
I have interests in many parts of computer science ranging from game, app, and web development to machine learning, embedded programming, and theory.
Skills
Languages
Frameworks
Other
Background and Passions
I’m a senior at MIT studying Computer Science (6-3) and Mathematics (18). I am passionate about creating tools and experiences for others to use, especially those that are technically challenging or interesting. There is nothing like building something useful for someone, or creating something that, under the hood, is special. When it comes to engineering for society, I am especially passionate about engineering towards a positive social impact. As a FLI, Dominican-American student, exploring how technology can play a role in academic outreach is particularly intriguing to me.
Extracurriculars
Learn more about my work on DormSoup as a member of SIPB.
More about my work with Code for Good.
A glimpse of the work I’ve done with MIT Motorsports.
August 2023 - May 2026
Background Information
Description
Each year, the MIT Motorsports team builds an electric racecar. As Telemetry/Data Acquisition Lead, I led the design and implementation of the vehicle telemetry system, which acquires data from the racecar and displays it in real-time. I developed dual-core STM32 firmware for CAN data acquisition, SD logging, and WiFi/serial streaming, and built a Next.js + FastAPI telemetry platform with parallelized Plotly visualizations and InfluxDB/QuestDB edge-cloud replication so team members can plot data from the car in real-time. The system contributed to a 10/10 design score from Formula SAE judges for the team’s LV/DAQ system.
Skills Developed
August 2024 - July 2024
Background Information
Description
Code for Good is a student organization @ MIT that pairs MIT students with local nonprofit organizations to provide advice and help with technical matters. During my semester with the club, my team and I worked with YouthBuild Boston, a local nonprofit organization focused on "providing under-resourced young people with the support and credentials needed to successfully enter the construction and design industry". The organization wanted a way to have their projects overlaid on a map that could show different statistics about neighborhoods in the Boston area and historical redlining practices. They also wanted to be able to update the map easily so we linked the website to a Google Sheet they could easily update.
Skills Developed
Website
August 2023 - May 2026
Background Information
Description
The Student Information Processing Board (SIPB), MIT’s first computer club, circa 1969, is "MIT’s volunteer student computing group, improving computing at MIT since 1969. Today, SIPB projects provide innovative services and special expertise to the MIT community." Some of SIPB’s projects are Hydrant, CourseRoad, and the project I work on, DormSoup. DormSoup is a web calendar service that utilizes LLMs to parse student event emails and displays them in a simple monthly view. Before my sophomore year, SIPB managed a project known as Dormdigest which had a similar purpose to DormSoup but didn’t utilize LLMs. In my sophomore year, DormSoup (which was being managed by other non-SIPB students) took Dormdigest’s place. I helped lead team meetings and contributed significantly to DormSoup’s switch from using OpenAI’s API to utilizing SIPB’s self-hosted open-source models. As of the end of my sophomore year, I officially took over as the project lead. As a SIPB keyholder and project lead, I maintain an Express.js/Postgres backend with vector deduplication and a searchable Next.js frontend serving 100 active users per month, along with a multi-stage, open-source LLM pipeline that detects, parses, and tags student event emails.
Skills Developed
Dormsoup
scrapTripLearn more about my first web development project.
Learn about some small ML projects I've built for fun.
More about the AAA style RPG I am working on.
June 2024 - August 2025
Background Information
Description
I've always thought that building a 3D game would be a very fun, educational, and challenging project. So, in the summer of my freshman year I began to work on creating a AAA style RPG with a friend of mine. While working on the game I develop skills with C++ and practice my creativity.
Skills Developed
scrapTripJanuary 2024
Background Information
Description
During web.lab, a month-long course and competition at MIT held between the fall and spring semesters, I collaborated with two teammates to develop scrapTrip—a web application that enables users to create virtual scrapbooks (“scraps”), connect with others, and generate travel itineraries using the Mobi API. This project marked my introduction to web development, providing hands-on experience as I built the site from the ground up.
Skills Developed
May 2025 - August 2025
Background Information
Description
After taking the Intro to Machine Learning course @ MIT, I wanted to do some small projects to discover topics not mentioned in class, deepen my understanding of some topics, and have some fun! So far, I've done small projects on classification with CNNs, image generation with GANs, and style transfer. I plan to continue to do more small, fun projects in my free time.
Skills Developed
Experiences
Experiences
Learn more about my software engineering internship at Nominal, building curve-fitting and benchmarking tools for engineering data.
Learn more about my work to develop data-driven and hybrid models to predict Area Control Error in power systems.
Learn more about my social impact work with the Cherokee Nation through the MIT PKG Public Service Center.
June 2026 - September 2026
Background Information
Description
During the summer before my senior year, I worked as a Software Engineering Intern at Nominal in New York, a platform that helps engineers analyze test and operational data. I built a custom curve-fitting feature that lets engineers fit, compare, and evaluate models against time-series data directly in Nominal. I also worked on the "Nominal for Motorsports" prototype, which supports motorsport telemetry analysis and race-engineering workflows, and built a benchmarking framework to measure performance, detect regressions, and guide optimizations.
June 2025 - August 2025
Background Information
Description
Code.Tulsa is an initiative by the MIT PKG Center that empowers MIT students to support marginalized communities in Tulsa through impactful internships with Native nations and local nonprofits. I collaborated with the Cherokee Nation, the largest tribe in the United States, representing over 450,000 citizens worldwide. My work focused on building an AI-powered legal search engine and assistant that uses FastAPI, React, and vector databases (BAAI/bge-large-en-v1.5 embeddings) to enable semantic search across Cherokee legal texts with automated source citation, helping lawmakers, judges, prosecutors, and other stakeholders efficiently research Cherokee law. I also developed web scrapers and data pipelines to extract, OCR, and structure unstandardized legal documents into JSON for downstream ML applications, and digitized the tribe’s registration process, further streamlining how individuals acquire Cherokee citizenship.
Skills Developed
June 2024 - September 2024
Background Information
Description
During the summer following my first year at MIT, I participated in my first UROP, building data-driven and hybrid LSTM models in PyTorch to predict Area Control Error (ACE) in power systems. The hybrid models combine physics-based forecasts with learned corrections. I evaluated forecast accuracy, data efficiency, and multi-step performance on real-world grid data, and co-authored a paper comparing mechanistic, data-driven, and hybrid approaches. The learned models demonstrated marginal improvements in certain metrics compared to the existing mechanistic models currently used for ACE forecasting.