About
Currently doing research on the privacy and personalization in FL
Work Experience
Invopop (YC 23)
Product Engineer
Instituto de Investigación Tecnológica
Research Assistant
Tower Consultores
Summer Intern
Universidad Pontificia de Comillas ICAI
Collaborator Student
Education
Georgia Institute of Technology
Universidad Pontificia de Comillas ICAI
Universidad Pontificia de Comillas ICAI
The University of Texas at Austin
Skills
Projects
The Effectiveness of Federated Learning in seismic interpretation
Presented at IMAGE26
sop-runtime: Reliable LLM Agents from YAML Procedures
YAML-defined procedures for LLM agents: code enforces identity checks, confirmations and data access; the model only reads and phrases. Guardrails, journey tests and pass^k evals.
Timeseries forecasting
Use of autoencoders and diffusion to forecast timeseries
Neural Network Library from Scratch [WIP]
Developed a bare C++ neural network library to learn about C++ and CUDA integrations
Diffusion Model from Scratch
Developed a diffusion model based on the UNet architecture from scratch to generate movie posters.
Federated Learning Framework
As part of my Capstone Project currently developing a Hierarchical Federated Learning framework based on the flower framework.
CFR-RL Generalization
Research on the use of Reinforcement Learning for Network Optimization. Based on the CFR-RL algorithm working on a model that is able to generalize to any topology.