_

Computer Science and Math @ IU '27 | SWE Intern @ Uber | Co-Founder @ Lawgic Prep

Gabe Shores

[ Projects ]

$ ls ~/projects10 itemstap to open
Lawgic Prep landing page

Lawgic Prep

  • Co-Founder and software engineer. Gamified LSAT practice and an official LSAC licensee, one of under 100 worldwide.
  • Built in React, JavaScript, Python, Google Cloud, and Stripe with an AI essay grader and 900K+ applications of admissions data.
  • Firebase backend serving over 6,000 questions across 58 tests with real-time analytics.
React JavaScript Firebase Python Google Cloud Stripe
Riley landing page

Riley

  • Won 1st place for Best Use of Browser Use at the YC Call My Agent Hackathon and took home AirPods Max.
  • One phone number for the whole family. Text it to check grades, manage calendars, approve spending, and relay messages.
  • Built in 12 hours with FastAPI, OpenAI tool calling, Browser Use cloud browsers, and Sponge payments on Solana.
Python FastAPI OpenAI Browser Use
Worldsmith terrain renders

Worldsmith

  • Natural language Minecraft terrain generator. Vanilla world generation reimplemented in Python, bit for bit.
  • An agent writes the density functions, renders the result, looks at the picture, and tunes the numbers.
  • Places whatever you describe on that ground, a hut, a lighthouse, a broken bridge or a castle, from a block grid with a shapes library for walls, towers, roofs and stairs.
  • Ships as a Claude Code skill with 482 passing tests, and installs the finished world into your saves folder ready to walk into.
Python Claude Code Minecraft
Limit order book and incremental update processing

Distributed Order Book

  • Dockerized Kafka to Spark pipeline reconstructing live order books at 50k+ events per second.
  • Concurrent Python ingestion normalizing Binance, Coinbase, and Kraken websockets into one schema.
  • Benchmarked on Jetstream2 with scaling sweeps and worker-kill fault injection. Judged best of roughly 30 projects in the class.
Kafka Spark Docker Python
PhysiCell diffusion solver research poster

PhysiCell Research

  • Contributing to the open-source PhysiCell project, a cell-simulation software.
  • Solved diffusion with numerical analysis in Python and C++, cutting complexity from O(n³) to O(n).
  • Parallelized routines with CUDA and OpenMP, speeding up large-scale simulations by over 300%.
C++ Python CUDA OpenMP
EduAdvisor AI course planning chat

EduAdvisor AI

  • Developed a full-stack, AI-powered course planning app using React, Tailwind, Vite, Flask, and the Gemini API.
  • Processed 100k+ records spanning over 8 years with NumPy and Pandas for more data-driven scheduling.
  • Students plan by chatting with it. Built with a team at a hackathon and deployed for anyone to try.
React Flask Gemini API Pandas
Social graph with one anomalous node flagged

Social Graph Anomaly Detector

  • ML pipeline flagging suspicious accounts with IsolationForest over PageRank and Louvain features.
  • Ingests live Reddit and X graphs, scores precision, recall and F1, and shows the result in an interactive PyVis app.
Python scikit-learn NetworkX PyVis
Luddy Scholars eclipse stream and outreach

Luddy Eclipse Project

  • Integrated OBS, NDI, and networking technologies to capture and stream live footage of the Great American Solar Eclipse, reaching over 3,000 online viewers and over 5,000 viewers at live events.
  • Collaborated with peers to engage over 200 elementary school students via outreach programs.
OBS NDI Networking
Neural network diagram

Java Neural Network

  • A neural network, perceptron, made fully in pure Java.
  • Able to score over 99% on MNIST digit recognition.
Java MNIST
Fintech Discord bot stock info message

Fintech Club Discord Bot

  • Built and deployed a Javascript and SQL app interfacing with Discord and trading REST APIs.
  • Developed dynamic and interactive bots allowing members to view and engage with the club's market fund.
JavaScript SQL Discord API

// hover a project to preview

[ Experience & Research ]

May 2026 – Aug 2026

Uber

Software Engineering Intern

  • Owned a system end to end tracking over one hundred reliability metrics, improving observability and catching regressions before they shipped.
  • Built Claude AI agent tools and integrations that streamlined developer workflows, cutting flagged issues by 90%.
  • Developed backend services on Fares, a Tier-0 system serving millions of riders and drivers, in Java, Python, and Go.
Java Python Go AI agents

May 2025 – Aug 2025

Allied Solutions

Software Engineering Intern

  • Engineered a serverless AWS document pipeline with Lambda, DynamoDB, and API Gateway, saving $64,000+ annually.
  • Modernized legacy C# .NET and SQL Server code, improving performance for a database serving 1,000+ banks.
AWS Lambda DynamoDB API Gateway C# .NET SQL Server

Nov 2024 – Present

Lawgic Prep

Software Engineer, Co-Founder

  • Secured a rare official LSAC license, one of under 100 worldwide, and negotiated pre-law sponsorships for early adoption.
  • Built an EdTech platform in React, JavaScript, Python, Google Cloud, and Stripe with an AI essay grader and 900K+ applications of admissions data.
  • Serve over 6,000 questions across 58 tests with real-time analytics on Firestore and Cloud Functions.
React JavaScript Firebase Python Google Cloud Stripe lawgicprep.org

Aug 2023 – Present

Indiana University Advising and Major Explorations Services

Peer Academic Coach, Data Analyst

  • Raised advising service usage using Tableau, Excel, and Python analysis, presented at a campus symposium.
  • Advise students one on one, helping them build class schedules, plan their degrees, and work through academic questions.
Tableau Excel Python

Aug 2025 – Present

LEMMA Research Lab

Undergraduate Researcher

  • Developed computational experiments in Python and C++ to study convex geometry and geometric probability.
  • Designed convex hull and point-inclusion algorithms using probabilistic theorems for geometric computation.
Python C++ Geometry

Aug 2024 – Dec 2025

Math Cancer Research Lab, PhysiCell Project

Undergraduate Researcher & Open-Source Contributor

  • Used numerical analysis in Python and C++ to solve diffusion, reducing complexity from O(n³) to O(n).
  • Parallelized routines with CUDA and OpenMP, speeding up large-scale simulations by over 300%.
C++ Python CUDA OpenMP

Leadership

Aperture Partners Vice President of Technology Jul 2026 – Present
IU Student Government Director of Technology May 2025 – Jun 2026
Fintech at Indiana University Director of Technology Aug 2024 – Present
"The struggle itself towards the heights is enough to fill a man's heart. One must imagine Sisyphus happy."

- Albert Camus

[ Get In Touch ]

Always open to discussing new projects, creative ideas, opportunities, or just to chat.