AI Engineer
I design and implement end-to-end AI systems, from classical machine learning to agentic systems, bringing together data, models, rigorous evaluation, and production software to solve real-world problems.
01
I am interested in the full path from an ambiguous problem to a system that can be evaluated, integrated, and operated. I am deliberately keeping my next step broad: I want to work on technically demanding ML and AI problems where sound evaluation and strong engineering matter.
02
Wollen Labs (now Luno)
April 2025 — April 2026
03
Team academic project — distributed product engineering
A social-location product spanning a mobile app, an administration backoffice, specialized backend services, deployment infrastructure, testing, security, and distributed observability.
Explore the GitHub organization04
Universidad de San Andrés
In progress
Coursework is expected to be completed in November 2026, with the thesis defense planned for February or March 2027.
Coursework combines mathematical and algorithmic foundations with classical machine learning, deep learning, computer vision, natural language processing, reinforcement learning, autonomous robotics, data systems, software engineering, cybersecurity, and responsible AI.
The program has been practical as well as theoretical: coursework and projects involved implementing and evaluating models, building data and training pipelines, and taking software systems through testing, integration, deployment, and observability.
Neural networks, CNNs, recurrent models, transformers, normalizing flows, diffusion models, multimodal models, self-supervised and contrastive learning, and multitask learning.
Classical ML, reinforcement learning, value- and policy-based methods, actor-critic architectures, probabilistic robotics, planning, and sensor fusion.
Databases, streaming, APIs, distributed systems, testing, CI/CD, Kubernetes, observability, security, fairness, privacy, and AI governance.
05
Papers and technical articles I co-authored.
A 2024 paper introducing convolutional KAN layers and evaluating them against conventional CNNs on Fashion-MNIST, including configurations with similar accuracy and roughly half the parameters.
Fairness criteria, causal reasoning, mitigation, governance, and monitoring across the ML lifecycle.
A practical comparison of ReAct, Plan-and-Execute, ReWOO, Tree-of-Thoughts, and multi-agent patterns.
Contact