Bryan Maristanez

TRIAGE AI - AI Triage Assistant

Developed an AI-powered triage assistant to optimize emergency room patient classification, reducing human bias and time-to-care in critical cases.

• Built a web app using Flask (Python) and MongoDB for real-time patient intake and severity classification.
• Integrated Gemini API via Agentic AI (Vellum) to simulate a medical triage expert capable of handling voice, image, and text inputs.
• Designed a dynamic 5-level triage system using Emergency Severity Index (ESI) metrics (e.g., vitals, consciousness, injury cause).
• Enabled transparent decision-making through AI-generated justifications to enhance clinical trust and oversight.
• Delivered a working UI prototype with Figma and a custom HTML/CSS/JS front end.

OPENDATA - Analyzing the correlation between apprehension types with socioeconomic factors from different Toronto Boroughs | YORK UNIVERSITY | April 2025 - April 2025
• Conducted both a research project and using OPENTORONTO data set.
• Merged multiple data sets to hypothesize topic, conducted data cleaning, data visualization, and statistical analysis.

• Presented my finding to OPENTORONTO judges and professors.

- tools: Google Collab, NumPy, Matplotlib, Seaborn, scikit-learn

ONE GUN - 2 Player shooter | Game development | May 2024 - June 2024

• Engineered a 2D multiplayer shooter incorporating distinct game mechanics such as interactive health bars, custom controls, multiple gameplay modes, and dynamic arenas, leveraging OOP principles for modularity and scalability.

Pomodoro Timer Web App
JavaScript, HTML, CSS, Node.js, Express
To learn the fundamentals of full-stack development, I built a Pomodoro timer app using JavaScript and a Node.js/Express backend

• Implemented timer logic (start, pause, reset) using setInterval() in vanilla JavaScript.
• Designed a simple and responsive UI with HTML and CSS.
• Created a lightweight Express backend to serve static files and practice server setup.
• Gained hands-on experience with client-server structure and time-based JavaScript functions.

Research: The surprising relationship with certain Methylation sites and it's predictability for future Alzheimers | STEM FELLOWSHIP | May 2025 - July 2025
• Implemented analytical models to investigate correlations between specific DNA methylation markers and the potential onset of Alzheimer's, enabling earlier risk

assessment through minimally invasive testing.

• Coursework: Introduction to Python, Calculus 1 & 2, Object Oriented Programming, Discrete Mathematics, Linear Algebra, Data Structures and Algorithms