Co-Founder & Head of Engineering

Juan Olloniego

Building GenAI products at massive scale for Latin America at Zapia AI

6M+

Users

$19.3M

Raised

10+

Years in AI

About

I'm a software engineer, startup founder, and former college professor with a deep passion for AI and its real-world applications. With an MSc in Artificial Intelligence from the University of Southampton (Distinction) and a background in software systems engineering, I've spent the last decade working at the intersection of research and production systems.

Currently, I'm the Co-Founder & Head of Engineering at Zapia AI, where we're building a GenAI-powered personal assistant for Latin America — serving 6M+ users and backed by Prosus Ventures, Endeavor Catalyst, and Factory HQ, with $19.3M raised.

Before Zapia, I worked on core ML systems at Landing AI (Andrew Ng's company), led client projects and pre-sales efforts at Tryolabs across computer vision, NLP, and recommendation systems, and taught postgraduate courses in Deep Learning and Reinforcement Learning at ORT University. I'm also an angel investor, backing startups like Lumepic and Viatik.

I thoroughly enjoy every step of the ML pipeline — from framing problems and validating ideas to shipping production systems that impact millions of users. When I'm not building, I'm probably on a tennis court, grinding ranked League of Legends, perfecting my neapolitan pizza dough, or chipping away at my ever-growing pile of research papers and non-fiction books.

Experience

Zapia AI

Mar 2023 — Present

Co-Founder & Head of Engineering

Building a GenAI-powered personal assistant for Latin America. Scaled to 6M+ users, raised $19.3M from Prosus Ventures, Endeavor Catalyst, Factory HQ, and Anthos Capital. Leading all backend engineering and infrastructure.

Landing AI

Nov 2021 — Mar 2023

Senior Machine Learning Engineer

Worked on core ML functionality for LandingLens. Designed and implemented large-image support for the platform, enabling new training and inference workflows across all clients.

Tryolabs

Oct 2019 — Nov 2021

Lead Machine Learning Engineer

Led client engagements, project delivery, and pre-sales across Computer Vision, NLP, tabular ML, and recommendation systems for US-based companies. Worked hands-on with GCP, AWS, PyTorch, OpenCV, Docker, and FastAPI.

ORT University

2016 — 2021

Professor — Postgraduate Courses

Taught Deep Learning (Bengio & Goodfellow curriculum), Reinforcement Learning (Sutton & Barto), Intelligent Agents, Foundations of Computing, and Programming Languages at both undergraduate and postgraduate levels.

Kaszek Ventures

Jan 2017 — Aug 2018

Product & Growth Analyst

Fulfilled internal tech needs and analyzed portfolio company data. Built and deployed an anomaly detection system for web traffic across all portfolio companies.

Education

University of Southampton, UK

2018 — 2019

MSc Artificial Intelligence

Graduated with Distinction (First Class). Focus on Machine Learning, Deep Learning, Computer Vision, and Reinforcement Learning. Dissertation on applying NLP techniques to code generation in Python.

ORT University, Uruguay

2013 — 2018

BSc Software Systems Engineering

Graduated with a cumulative grade of 87/100. Courses in software architecture, data structures, algorithms, theoretical CS, and machine learning.

Projects

Paper Implementations

Re-implementing research papers across different ML domains with a focus on simplicity and reproducibility. All implementations are limited to a Colab runtime environment for easy access.

PyTorchDeep LearningResearch

Deep Learning Notebooks

Collection of teaching notebooks covering deep learning topics from computer vision to NLP, developed for postgraduate courses at ORT University.

JupyterTeachingPyTorch

Deep Learning for Code Completion

MSc dissertation project applying NLP techniques to code generation in Python. Explored line-level and block-level completion using deep learning, comparing models trained on different tasks.

NLPCode GenerationMSc Project

Few-Shot Optimization

Reproduction of "Optimization as a Model for Few-Shot Learning" (ICLR 2017). Group project implementing novel neural network architectures for meta-learning.

Meta-LearningResearchICLR

Skills

AI & ML

GenAI / LLMsDeep LearningNLPComputer VisionReinforcement LearningML Systems Design

Languages & Frameworks

PythonPyTorchScikit-LearnFastAPINumPyPandasOpenCVTypeScript

Infrastructure

GCPAWSDockerLinuxGitCI/CD

Other

System ArchitectureTeam LeadershipTechnical HiringTeachingSQLHaskell

Contact

Have something in mind? I'd love to hear from you.

© 2026 Juan Olloniego