Projects

Scaling GPU, Compute, and Storage infrastructure across 400+ Azure datacenters worldwide.

• Validating and securing hardware SKUs from factories to customers for cloud infrastructure
• Delivered a security initiative for secure hardware onboarding — preventing devices from accessing other devices until fully provisioned
• Promoted in under 1 year — rare for a graduate hire
• Also performing PM and team leadership work within role

Built sensor modules, collected real-world data, and trained ML models to predict campus e-jeepney arrival times with near-perfect accuracy.

• Co-authors: Anna Ysabella Habana, Franco Manuel D. Velasco
• Advisers: Patricia Angela R. Abu, PhD; Raphael B. Alampay, PhD
• Accepted for Oral Presentation at Macau
• Built 3 Raspberry Pi sensor modules (GPS, temperature, humidity) placed on campus e-jeeps across 9 operational days
• Full ML pipeline: data collection → 42 feature extraction → preprocessing → model training → hyperparameter tuning (Bayesian Optimization) → feature reduction (PCA + Autoencoders)
• Transformer + Autoencoders achieved R² = 0.9992 — near-perfect prediction
• Demonstrated Transformers’ viability for transportation time-series data

Built an RFID-based attendance system from scratch using Arduino, a spare PC, GSM module, and RFID cards.

• Features: attendance logging, parent SMS notification, teacher email summaries, visual interface
• Tested with 100 students, 100 parents, 5 teachers
• Statistically significant improvement in adoption and positive perception
• Won Best Thesis in Applied Sciences & Engineering

Built "Refind": an app bridging the preloved clothes market into a dating-swipe format.

• Built the team, developed the idea, and pitched — all in 2 days

Proposed CISO recommendations for IAM operations globally by analysing automation failures.

• Built a tool independently that saves 8 hours/month from analysts globally
• Led as main local developer migrating manual work into AI virtual assistant (Kore.AI)
• Received return offer for exclusive Management Trainee graduate program

Smart agriculture advisory app using weather and IoT data to advise farmers on optimal crop strains.

• Incorporated farmer knowledge feedback loops and collective coordination to prevent market imbalance