Projects on C++, Rust, Node, Python, AI/ML

This is Redact, a prototype for high-performance redaction, ideal for legal teams, compliance departments, and any organization handling sensitive document processing. This involved building an optimized engine in C++ to automate the protection of sensitive data within massive document sets. This includes optimized string manipulation and streaming Input-Output for near-zero memory overhead. It aims to protect client confidentiality by ensuring precision in redactions.

I continued the security-focused engine for Redact, focusing on memory safety and data integrity. This involved leveraging Rust's ownership model to prevent unauthorized memory access during file processing. This includes the implementation of digital hashing to ensure a secure chain of custody. It aims to provide a tamper-proof digital fingerprint for sensitive documents.

Responsible for back-end and front-end software engineering relying on Node.js ecosystem, and integration tasks for one of Philippines’ most trusted and widely used payment platforms. Today, it powers millions of secure transactions for top local and global brands such as TikTok, Lazada, Starbucks, SM, AirAsia, and Converge ICT.

I was an AI Intern for Philippines' Department of Science and Technology - Advanced Science and Technology Institute, where I contributed to Project ROAMER (Robot for Optimized and Autonomous Mission-Enhancement Response). This involved developing a smart robot capable of detecting banana plant diseases. This includes in-house data labeling, neural network training, the science of making C++ and Python work together, and deployment in production running on our AI supercomputer. Today, it aims to advance agricultural diagnostics, supporting farmers with early disease detection for healthier crops.

I led the development for 'Video Lens,' an award-winning AI thesis. This involved understanding first how C powers the Python language used, and understanding C++ that powers the 5-stage AI models in Python for large-scale video analysis. This includes transcription, summarization, and NLP-driven querying, and the science of deploying it in production via Docker and AWS. Today, it helps users transform hundreds of hours of video into searchable, actionable data.