I’m an AI researcher at Sailplane, where I work on developing AI agents for software engineering. My research and development focuses on improving agents’ ability to understand developers’ intent and to learn to perform complex tasks with and without fine-tuning.
Previously, I was a postdoctoral researcher associate at ESAT-PSI, KU Leuven, working under the guidance of Tinne Tuytelaars and Marie-Francine Moens on continual learning and parameter-efficient fine-tuning for language models.
I completed my joint PhD at KU Leuven and Pontificia Universidad Católica de Chile under the supervision of Professors Alvaro Soto and Marie-Francine Moens, focusing on improving language models by integrating insights from human language processing. Throughout my academic career, I was associated with several research laboratories, including LIIR, IALab, IMFD, and CENIA.
As someone originally from Quito, Ecuador, I’m especially committed to advancing NLP for Spanish. I’ve contributed to the development of open Spanish language models such as ALBETO, DistilBETO, BARTO and T5S. I’ve also been involved in creating several datasets, including SemRel2024, CVQA, BRIGHTER. These efforts aim to support accessible, high-quality resources for research and applications in Spanish NLP.
Recent News
- May, 2025. Our paper “BRIGHTER: BRIdging the Gap in Human-Annotated Textual Emotion Recognition Datasets for 28 Languages” has been accepted at ACL 2025.
- Sept, 2024. Our paper “Learning to Route for Dynamic Adapter Composition in Continual Learning with Language Models” has been accepted at EMNLP 2024.
- Sept, 2024. Our paper “CVQA: Culturally-diverse Multilingual Visual Question Answering Benchmark” has been accepted at NeurIPS 2024.
Research Interests
- Natural language processing
- Large laguage models (LLMs)
- In-context learning
- Reinforcement learning
- Continual learning
- Recommender systems