Available for opportunities

Currently: SSC President @ TIP Manila

Dexter
Soriano

BS Computer Engineering · Data Science

Building computer vision systems and data-driven applications at TIP Manila.

Request CV View Projects

01 — About Me

Data Science Student
& Council President

I'm a Computer Engineering student majoring in Data Science at T.I.P. Manila, with a deep interest in machine learning, computer vision, and building intelligent applications. As the current SSC President, I combine technical problem-solving with leadership — aiming to create meaningful impact through data-driven technology and continuous learning.

View Experience

02 — Technical Toolkit

Languages & Tools

HTML

Semantic markup & structure

<!DOCTYPE html>
<html lang="en">
  <head><title>Sample</title></head>
  <body>
    <h1>Hello, World!</h1>
  </body>
</html>

CSS

Styles & layouts

body {
  background: #f9f9f9;
  font-family: 'Arial';
}
h1 {
  color: #0055aa;
}

JavaScript

Dynamic interactions

window.addEventListener(
  'resize', () => {
  const w = window.innerWidth;
  const h = window.innerHeight;
});

React

Component-based UI

function ProjectCard() {
  return (
    <article className="card">
      <h3>React Project</h3>
    </article>
  );
}

PyTorch

Deep learning & training

import torch
model = YOLO('yolov8s.pt')
x = torch.rand(1, 3, 640, 640)
out = model.predict(x)

Python

Data science & ML

from ultralytics import YOLO
model = YOLO('yolov8n.pt')
results = model.train(
  data='data.yaml',
  epochs=100, imgsz=960)

Pandas & NumPy

Data wrangling & analysis

import pandas as pd
df = pd.read_csv('data.csv')
df.dropna(inplace=True)
print(df.describe())

SQL

Queries & databases

SELECT name,
  AVG(score) AS avg
FROM results
GROUP BY name
ORDER BY avg DESC;

OpenCV

Computer vision

import cv2
cap = cv2.VideoCapture(0)
circles = cv2.HoughCircles(
  gray, cv2.HOUGH_GRADIENT,
  dp=1.2, minDist=40)

Snapshot — 2026

Impact Highlights

Lead Role

SSC President

TIP Manila · 2025 to 2026

Projects Built

11+

Data science, fullstack, and frontend

Core Stack

ML + Web

Python, YOLOv8, OpenCV, JS

Current Focus

Computer Vision

Real-time inference and product UX

03 — What I've Built

Featured Projects

Showing projects

01 Data Science

Coin-Scan

Real-time Philippine coin detector and counter using a YOLOv8 ensemble (nano + small models). Features OpenCV circle detection, size-based auto-calibration, and dual inference — a browser WebAssembly UI and a Python CLI. Classifies ₱1, ₱5, ₱10, ₱20, and arcade tokens with live PHP value totaling.

YOLOv8 PyTorch OpenCV ONNX Python

Source Code

02 Fullstack

RecipeChecker

A fullstack recipe management system built as a group project. Features a relational SQL database backend, normalized schema design, and a Python GUI — covering CRUD operations, search, and user-facing browsing.

Python SQL DBMS Fullstack

Source Code Live

Portfolio Website

A fully custom-designed personal portfolio built from scratch — particle canvas, GSAP scroll animations, glassmorphism cards, and a fully responsive layout with zero dependencies.

HTML CSS JavaScript GSAP

Source Code Live

04 Data Science

Match Prediction

League of Legends match outcome predictor trained on large-scale ranked game data. Parses champion, item, summoner, and team tables to build ML models that forecast win probability before a match ends.

Python Pandas scikit-learn ML

Source Code

05 Data Science

NeuralLab

Neural network lab notebook implementing feedforward architectures for multi-class classification on the Iris dataset. Covers forward propagation, backpropagation, and activation function experiments.

Python Neural Networks Jupyter

Source Code

06 Data Science

Chip

Exploratory data analysis on a semiconductor chip dataset. Investigates CPU/GPU performance characteristics, identifies trends across chip generations, and extracts insights through statistical analysis and visualization.

Python Pandas Jupyter

Source Code

07 Data Science

Neural

Linear regression on the Ames Housing dataset for house price prediction. Features feature engineering, model training, and both a Jupyter notebook walkthrough and a standalone Python implementation.

Python scikit-learn Jupyter

Source Code

08 Data Science

Classification

Multi-model classification study on the Iris dataset. Benchmarks Decision Tree, K-Nearest Neighbors, Logistic Regression, and Random Forest classifiers — with confusion matrices and accuracy comparisons.

Python scikit-learn Jupyter

Source Code

09 Data Science

Titanic

Kaggle Titanic survival prediction challenge. Performs EDA on passenger demographics, engineers features, and trains classification models — visualized through feature importance and model comparison plots.

Python Pandas scikit-learn Jupyter

Source Code

10 Data Science

Regression Models

House price regression on the Ames Housing dataset. Compares multiple regression models with actual vs. predicted plots, sale price log-transformation analysis, and comprehensive error metrics.

Python scikit-learn NumPy

Source Code

11 Fullstack

Project UNITE

A deployed fullstack web application built with a Django backend and a clean, product-style frontend experience. Designed for accessible interface patterns and a practical real-world deployment workflow.

Django HTML CSS JavaScript Fullstack

Source Code Live

More on GitHub

Explore all my repositories, contributions, and ongoing experiments on GitHub.

View Profile

Explore Live

Interact with selected projects — browse notebooks, live apps, and repos below.

Data Science

Coin-Scan

Real-time Philippine coin detector and counter using a YOLOv8 ensemble (nano + small models). Features OpenCV circle detection, size-based auto-calibration, and dual inference with browser WebAssembly UI and Python CLI. Classifies ₱1, ₱5, ₱10, ₱20, and arcade tokens.

Open Repo Open Repository

04 — Leadership & Involvement