Stelier
Software Engineer Team Lead
Jun 2026 - Present
Leading a three-person team building a cross-platform marketplace for handcrafted goods and services with React Native, Expo, Supabase, and PostgreSQL.
Kyle Yumul
Aspiring Software Engineer
CS |
About Me
I am a Computer Science student at the University of Florida, graduating in May 2028 and pursuing an AI Applications Certificate. I build full-stack web and mobile applications and enjoy turning data and artificial intelligence into useful products.
What really led me to my interest in this field was my constant involvement with technology as a kid and seeing the variety of different things I could do with it and what it could do for me. From then on, since before middle school, I wanted to be the one behind the scenes making the technology that people use every day.
Today, I lead a three-person engineering team at Stelier and an eight-person team building the UF SASE mobile app. My recent work includes a software engineering internship at BeachLens, a restaurant dish-rating app called Whatchu Get?, and Ball Knowledge, a soccer analytics and AI project that brings together my interests in software and the game.
Recent Experience
UF SASE
Visit UF SASESoftware Engineer Team Lead
Sep 2025 - Present
Leading an eight-person mobile app team and improving events and attendance workflows for a community of 500+ members.
BeachLens
Visit BeachLensSoftware Engineer Intern
Apr - Jul 2026
Built fast beach-discovery interactions and matched 100+ beaches to NOAA and FWC environmental sampling stations using Next.js and Supabase.
Citibank
Software Engineer Mentorship
Mar - Apr 2026
Selected from 4,000+ applicants for Citi’s Early Identification Technology Program, with mentorship in software infrastructure and agile engineering.
Tech Stack
My recent projects span cross-platform mobile development, cloud data pipelines, and AI applications.
Languages
Frameworks
Developer Tools
Also Working With
- React Native
- Expo
- SQL
- PostgreSQL
- pgvector
- AWS S3 & Lambda
- scikit-learn
- NumPy
- Streamlit
- Groq
- Drizzle ORM
- Turso
- Redis
Projects
Updated October 15, 2026
Soccer · Data · AIBall Knowledge
128matches462K+events8,940evidence chunks
Ball Knowledge (Sep 2026 - Present)
A soccer match-intelligence app built with Python, FastAPI, AWS, PostgreSQL, and pgvector. It processes 462,000+ events across 128 matches, estimates expected goals with scikit-learn, and answers match questions with cited evidence using Groq. The latest local pipeline indexes 8,940 evidence chunks across the full match catalog.
Slide 1 of 5: Ball Knowledge (Sep 2026 - Present)