Skip to main content

Roberto Villafuerte

AI Engineer · MLOps · RAG Systems

Download PDF (opens in new tab)

Summary

Final-year Computer Science undergraduate specializing in MLOps and retrieval-augmented generation (RAG) systems. Experience architecting automated forecasting pipelines and deploying multi-tenant GenAI support agents. Completed Master's-level coursework in Trustworthy ML at the University of Helsinki. Focused on shipping scalable, reliable, and bias-aware production AI systems.

Experience

AI Engineer

COMPUMAX · Part-time, Remote

Dec 2025 – Present
  • Architecting a multi-tenant RAG helpdesk for small businesses
  • Designing end-to-end backend in Python, from prototype to deployment
  • Building internal GenAI workflows aligned to business goals and scalable architecture

Machine Learning Engineer Intern

Datalysis Group · Remote

Feb 2025 – Jun 2025
  • Automated demand forecasting pipeline for inventory replenishment
  • Implemented end-to-end workflow: cleaning/validation, feature engineering, training, evaluation
  • Stack: Snowflake, Mage AI, Python

Education

B.Eng. Computer Science

Universidad San Francisco de Quito (USFQ)

2022 – 2026

Focus: software engineering, algorithms, databases, applied machine learning. Expected graduation: Dec 2026.

Exchange Studies — Data Science (Master's-level)

University of Helsinki

Aug 2025 – Jan 2026

30 ECTS · GPA: 4.83/5.0

Coursework: Big Data Platforms, Computer Vision, Engineering of ML Systems, Trustworthy ML, Software Architecture.

Certifications

Technical AI Safety Course

BlueDot Impact

Jan 2026

General Data Protection Regulation (GDPR)

Packt

Dec 2025

IBM Machine Learning Specialization

Coursera

Awards

2nd Place — IEEEXtreme 18.0 (Ecuador)

Global 24-hour algorithmic programming competition organized by IEEE

Oct 2024

Skills

Languages

Python SQL Bash

AI / ML

Generative AI LLMs RAG LLM Evaluation Forecasting ML Systems

Infrastructure & Engineering

Snowflake Mage AI Docker CI/CD GitHub Actions Git

Concepts

MLOps Data Pipelines Reproducibility Validation CRISP-DM