Hi, my name is

Engineering Portfolio

Joshua Ta
01.

About Me

I'm a Computer Science and Economics student at Brown University excited about integrating AI processes for medtech applications. I have a background in competitive robotics and can nerd out about the robotics and computer building. I love building systems that leverage new technologies to solve real problems and create impactful products.

My experience spans full-stack development, machine learning research, and hardware engineering. Currently, I'm working with Professor Rafael La Porta on using LLMs to extract analysis from analyst reports to build financial models, while also developing various personal projects like web apps, video games, and MedTech solutions!

02.

Featured Projects

ModMed AI Agent

ModMed AI Agent

HIPAA-compliant ChatGPT-style assistant for streamlined patient data access. Integrated LLaMA model with FastAPI backend and OAuth2 authentication for real-time clinical information summarization.

React FastAPI LLaMA Docker OAuth2
→ https://uroassist.net/login
Cad4Kids

Cad4Kids

GUI application converting 2D images into interactive 3D models exportable as STL files. Optimized rendering with vectorized NumPy operations for improved performance and memory efficiency.

Python PyQt5 OpenGL NumPy 3D Printing
VEX Robotics

VEX Robotics Champion

Founded and led globally competitive robotics team to regional championship. Implemented PID control and odometry-based motion planning for autonomous navigation, placing in top 0.3% globally.

C++ Arduino PID Control Motion Planning
→ https://www.youtube.com/@2922_Charles
LLM Financial Analysis

LLM Financial Analysis

Research project extracting and structuring data from 120,000+ analyst reports using Llama 3.1 and FinBERT. Analyzing left-tail market expectations from Morgan Stanley and JP Morgan forecasts.

Python Llama 3.1 FinBERT HPC NLP
Crypto Mining

Crypto Mining Operation

Engineered and optimized mining computers achieving 150 MH/s from RTX 30 Series. Invested $7,000 to generate $25,000 profit, reinvested in diversified portfolio including ETFs and index funds.

Hardware Optimization Investment Analytics
Coding Campers

Coding Campers

Dragon Kim Foundation Grant winner ($5,000). Provided free coding education to elementary and middle school students, teaching Python, Scratch, and binary concepts through interactive workshops.

Python Scratch Education Leadership
03.

Skills

Languages

Python
JavaScript / TypeScript
Java
C/C++
HTML & CSS
SQL / STATA

Frameworks & Tools

React
FastAPI / Flask
Node.js / Next.js
Docker
Git
TailwindCSS

AI & Data

LLaMA / FinBERT
NLP
NumPy / Pandas
OpenGL
HPC Clusters
Data Analysis
04.

Experience

February 2026 – Present
Product Manager
Full Stack at Brown

Promoted from developer to PM; leading 10+ developers and designers to build full-stack web apps for clients. Coordinating design and engineering workflows in Figma/Git and Next.js/React, managing full product lifecycle. Implemented a CMS with Cosmic enabling client writers to publish, tag, and organize multimedia content.

January 2025 – Present
Research Assistant
Brown Economics Department

Deployed Llama and FinBERT pipelines on Brown's CCV cluster to extract data from 120,000+ analyst reports. Automated preprocessing with Selenium and PyMuPDF, reducing data scraping and cleaning time by 85%. Developed STATA scripts for regression analysis on market expectations and built Python matching scripts to link IBES analyst IDs to Refinitiv codes.

June 2024 – August 2024
Data Analytics Intern
University of Southern California

Improved rendering speed of 2D to 3D array transformation by 40% with optimized MATLAB/Python algorithms. Built an interactive GUI with OpenGL providing advanced CAD tool accessibility and STL visualization. Collaborated with graduate researchers using Agile workflows and weekly standups.

May 2023 – August 2024
Founder / Co-Captain / Programmer
VEX Robotics Team #2922C

Founded and scaled a VEX robotics team to top 0.3% globally, recruiting 8 members and securing $7,000 in funding. Programmed PID control and odometry motion planning, achieving 2.1x autonomous speed and reducing error by 60%. Won regional championship and OCSA Programming Award.

05.

Get In Touch

joshua_ta@brown.edu GitHub LinkedIn