Marvin
Mangubat

AI Systems Engineer & Computational Architect

Executing a disciplined roadmap toward advanced computational integration. Architecting cross-domain AI platforms (with a computational biology focus) and deploying scalable LLM analytics to build the technological foundation of the future.

Systems Architect
COMP.INFRA ACTIVE

About

I am an AI developer and software engineer specializing in systems engineering and computational infrastructure. I work on projects ranging from full-stack application development and process optimization to advanced bioinformatics solutions.

I have directed engineering teams to deliver centralized management systems for non-profits and built scalable Python-LLM analytics pipelines that drastically cut review times for multinational tech platforms. Alongside technical development, I have also shaped institutional standards by authoring comprehensive policies on the ethical use of Generative AI.

Lately, I have been diving deeper into AI-driven bioinformatics, specifically leveraging protein language models for generative design. I treat this complex infrastructure as a continuous sandbox, where my daily structure prioritizes rhythm over intensity, and every friction point is captured as training data to compound long-term results.

Inspect Full Resume

Technical Capabilities

Core Domains

  • Artificial Intelligence
  • Machine Learning
  • Bioinformatics
  • Systems Design
  • Cross-Functional Leadership

Infrastructure

  • Python & Jupyter
  • Node.js & Express
  • MongoDB & SQL
  • Java
  • C

Interfaces & Tools

  • React & JavaScript
  • Software Testing
  • Git Version Control

The Trajectory

BSMS Computer Science

2025 — 2028

De La Salle University - Manila

Executing the rigorous BS/MS track with Magna Cum Laude standing. Maintaining a core focus on Artificial Intelligence, Machine Learning, and Theory of Computation as a foundation for advanced bioinformatics architecture.

Strategic Industry Integration

2025

Shopee • VertexPro • Unbound Manila

Transitioned theoretical models into high-leverage production environments. Engineered automated Python-LLM data pipelines, deployed secure MERN-stack architectures, and directed cross-functional engineering teams as a Scrum Master to resolve core operational bottlenecks.

Organizational Leadership & Policy

2024 — 2025

DLSU Legislative Assembly • Computer Studies Government • SPRINT • LSCS

Authored and implemented the comprehensive ethical Generative AI policy as a Batch Legislator, standardizing academic integrity across the university. Scaled external partnerships and industry engagements as the Director for Relations, AVP for Externals, and AVP for University Relations to bridge the gap between academic theory and industry execution.

Recent Projects

Full Archive

PhageLytic-ESM Architecture

Machine Learning • Conference Paper

2026

Developed a machine learning pipeline to classify phage lytic proteins for therapeutic applications, converting unstructured biological data into vectorized inputs. Engineered high-dimensional embeddings from ~2,500+ enzyme sequences using the PhaLP 2.0 database and the 565 MB ESM-2 model.

ESM-2 Jupyter Machine Learning

Generative Phage Lysin Architecture

Master's Thesis • AI Validation

2025 - Present

Developing a hybrid Generative AI pipeline utilizing ProGen2 and ESM-2 to engineer novel phage lysins targeting multi-drug resistant Klebsiella pneumoniae. Planning to implement computational validation workflows integrating AlphaFold2 for structure prediction, Foldseek for structural clustering, and GROMACS for molecular dynamics simulations.

ProGen2 AlphaFold2 GROMACS

Operations Analytics Pipeline

Shopee • Business Efficiency

2025

Architected a scalable LLM analytics engine to synthesize customer-agent transcripts, cutting review time by over 95%. Developed automated data reconciliation across 15+ disparate datasets, creating a defensible audit trail for executive reporting and enabling real-time operational visibility.

Python LLM Pandas

Centralized Case Management System

Unbound Manila • Systems Design

2025

Directed a cross-functional development team of 7 engineers to architect and deliver a centralized system serving ~100 users. Managed sprint life cycles, translated partner requirements into actionable technical sprints, and integrated iterative systems design improvements.

MERN Stack Cypress Scrum