LOUDY
MIGUEL.
Let’s talk ↗

Loudy Miguel TorrejasEntrepreneur and Founder of GenXYZBig ideas.Real-worldimpact.

From concept
to something real.
AI, applications and connected machines, built for your next big move.

Full-body portrait of Loudy Miguel Torrejas, founder of GenXYZ, in a black business suit with arms openGenXYZ Lab code editor on a tabletTime Machine app on a phoneMarket Flow interface on a tablet

I turn ambitious ideas into useful products, for entrepreneurs, for teams, and for what’s next.

SCROLLBASED IN THE PHILIPPINES
BUILDING FOR EVERYWHERE
AI ENGINEERING✳FULL-STACK SYSTEMS✳ROBOTICS & IoT✳SEO & QA✳FOUNDER OF GENXYZ✳
01 / SELECTED WORK

Less talk.
More built.

From useful apps to autonomous machines: nine builds across software, AI, and hardware. Different challenges, one drive to make things work.

01 / LEARN. PRACTICE. BUILD.GenXYZ Lab code practice interface, built by Loudy Miguel Torrejas
FLAGSHIP OF GENXYZ / WEB & MOBILE

GenXYZ Lab

The flagship learning platform of GenXYZ, the company I founded. Offline-first quizzes, course content, and a hands-on code practice workspace that keep learning moving, even without a connection.

↗
02 / MAKE TIME YOURS.Time Machine alarm app in a phone mockup
PRODUCTIVITY / MOBILE APP

Time Machine

Alarms, timers, a stopwatch, and a chess clock in one app. Everyday tools for more intentional time.

↓
03 / INTELLIGENCE IN MOTION.Robot Controller recycling robot logo
ROBOTICS / COMPUTER VISION

Robot Controller

A mobile command center for TrashBot: robot navigation, drive and arm controls, and waste collection monitoring.

95.15% mAP@5090.85% PrecisionValidation · epoch 100
04 / EMBEDDED SYSTEMS / HARDWAREHand-built ESP32 Super Calculator electronics
EMBEDDED SYSTEMS / HARDWARE

Super Calculator

A custom calculator built around ESP hardware and Arduino, bringing software logic into a compact physical device.

  • ESP
  • Arduino
05 / ROBOTICS / AUTONOMOUS SYSTEMSTrash Rover autonomous waste collection robot with a robotic arm
ROBOTICS / AUTONOMOUS SYSTEMS

Autonomous Trash-Picking Robot

An autonomous waste-collection robot combining computer vision, navigation, and a robotic arm to detect and pick up litter.

  • Docker
  • Raspberry Pi
  • ROS 2
  • YOLO
  • Python
  • C++
  • Arduino
  • Computer Vision
06 / COMPUTER VISION / WEB APPLICATIONPaper Checker interface for camera-assisted answer-sheet review
COMPUTER VISION / WEB APPLICATION

Paper Checker

Camera-assisted test-paper review: capture answer keys, scan student sheets, and bring scoring and reporting into one workflow.

  • Flask
  • MySQL
  • Python
  • Computer Vision
07 / COMMERCE / FULL-STACK APPLICATIONMarket Flow communities and marketplace interface
COMMERCE / FULL-STACK APPLICATION

Market Flow

A connected commerce app that brings communities, stores, and everyday business tools into one workspace.

  • Redis
  • MySQL
  • Flutter
  • Go
  • Cloudflare
  • Firebase
08 / MACHINE LEARNING / PROTOTYPEBanknote authenticity prototype showing manually entered variance, skewness, and entropy features
MACHINE LEARNING / MANUAL FEATURE INPUT

Banknote Authenticity Detection

A machine-learning prototype that classifies banknotes as genuine or counterfeit using manually entered numerical features. Designed with computer vision in mind; this version uses input fields rather than camera-based feature extraction.

  • Machine Learning
  • Classification
  • Manual Feature Input
09 / COMPUTER VISION / DESKTOP APPLICATIONAttendance Checker class analytics and student attendance dashboard with anonymized sample data
FACE RECOGNITION / ATTENDANCE MANAGEMENT

Attendance Checker

A Java desktop application combining face detection and recognition with attendance tracking. Class analytics, individual attendance summaries, and session history make records easier to review.

  • JavaCV
  • Machine Learning
  • Face Detection
  • Face Recognition
  • Java
  • Maven
  • NetBeans

Preview uses anonymized sample data.

02 / WHAT I CAN BUILD FOR YOU

Your ambition.
My engineering.

From a founder’s first launch to a company’s next system. I connect design, software, and hardware to solve the whole problem.

IDEA → FIRST LAUNCH<7DAYS

Small scope.
Serious momentum.

Focused websites and app MVPs in under seven days, with scope agreed upfront. Larger systems follow a tailored roadmap.

Define your sprint ↗
01

Full-stack websites & apps

Customer-facing websites, mobile apps, dashboards, and the backend systems that power them.

02

AI that does useful work

Intelligent features, workflow automation, and computer vision built around a clear business need.

03

Robotics & connected devices

IoT prototypes, remote control interfaces, and hardware–software integration that brings ideas into the physical world.

04

Findable. Reliable. Ready.

Technical SEO, responsive interfaces, and quality assurance to help your product reach people and work for them.

03 / THE PERSON BEHIND THE BUILD

Curiosity starts it.
Craft finishes it.

I’m Loudy Miguel Torrejas, an entrepreneur and the founder of GenXYZ, turning ideas into products across software and the physical world.

I work with entrepreneurs and companies to turn complex ideas into practical products. From the first sketch to the final quality check, I care about how things look, how they work, and what they make possible.

01
Define
02
Build
03
Test
04
Launch
04 / THE COMMON ROOM

Good ideas
start with hello.

Ask about a project or share what you’re building. A public space for curious people: keep it kind, useful, and free of personal information.

THE BUILDERS’ ROOMNot connected

A little space for big conversations.

The community is opening soon. In the meantime, say hello using the contact form below.

Messages are public. No links, contact details, or spam. One post per minute.

05 / YOUR NEXT BIG THING

Let’s make
it happen.

Tell me what you’re building, who it’s for, and what success looks like.

Sent privately to my inbox. Your email won’t appear in the community.

TRASHBOT / VALIDATION RESULTS

A closer look
at the model.

Object detection validation metrics from the supplied TrashBot v1 training run, at epoch 100. These are validation results, not a claim of real-world accuracy.

95.15%mAP@5077.47%mAP@50–9590.85%Precision89.22%Recall
Training and validation loss, precision, recall and mAP curves over 100 epochs

The app documentation describes a classification model; this supplied run contains object detection metrics. They are presented as a separate training experiment. No download is provided.