Experience

2026 – present Doodle AG · Zurich, Switzerland

Highlights

Designed and evaluated 20+ event-importance features across three scheduling products, including an event-topic classifier using multilingual sentence embeddings in Python and SQL.
Built a two-stage multilingual event-topic pipeline: derived a topic taxonomy from historical data using BERTopic, assigned events to it using sentence embeddings, and validated outputs through structured manual review.
Engineered 6 dbt models on Amazon Redshift to compute meeting- and attendee-level features at a production data-warehouse scale, with comparative feature analysis on representative samples.
Implemented end-to-end Apache Airflow DAGs connecting the data models to the event classification step.

2023 – 2025 Idiap Research Institute · Martigny, Switzerland

Highlights

Led the SEM24 Innosuisse project, transforming business requirements into deployed NLP systems for skill extraction from job offers and resumes through effective academia-industry collaboration.
Delivered three multilingual, explainable skill-extraction systems for HR recommendation: rule-based, semantic search, and a fine-tuned BERT model applying Named Entity Recognition (NER), text classification, and prompt engineering in PyTorch and Hugging Face, with the BERT system adopted by the industry partner.
Published research findings across 5 peer-reviewed papers (RecSys 2024, HICSS 2025, SwissText 2025–2026, Data & Policy 2026) and delivered technical talks at international NLP conferences.
Supervised a three-month internship applying NLP to social media data to analyse public perspectives on climate change.

Highlights

Conducted research on document-level text simplification, including statistical analysis, generative AI models, and evaluation metrics design, resulting in 3 publications (ACL 2021, SEPLN 2021, RANLP 2023).
Developed methods and benchmarks for lexical and sentence simplification using LLM prompt engineering, publishing 2 papers (EMNLP 2022, EMNLP 2023).
Implemented multilingual methods for readability research, resulting in a publication (EMNLP 2022).

2011 – 2020 Intel Corporation · Heredia, Costa Rica

Highlights

Internal Tooling & Data Automation: Developed data processing and analysis platforms, continuous integration pipelines, and code generation libraries for engineering teams to validate microprocessor designs before production, reducing manual engineering hours in Python, Perl, Java, and Ruby (9 years).
Product Ownership & Team Lead: Led a 5-person automation engineering team, aligning technical requirements with business goals (1 year); coached cross-functional teams as Scrum Master (3 years).

Leadership & Service

August 2025 Idiap Create Challenge (ICC) · Martigny, Switzerland

Led and co-developed HRme, an AI-driven behavioural screening platform using conversational LLMs and speech-to-text (ASR) to automate interview evaluation via STAR scoring and O*NET competency mapping to reduce hiring bias.

2022 – 2026 NLP conferences and journals

Reviewed for NLP4Sustain (2025), ClimateNLP (2024–2025), ACL ARR (2024–2026), BioLaySumm (2023–2024), Springer LRE (2023), and the TSAR Workshop (2022–2024).

February 2024 Idiap Research Institute · Martigny, Switzerland

2020 – 2022 University of Manchester · Manchester, United Kingdom

Facilitated workshops, guided students in problem-solving, and supported online discussions. Evaluated student work, provided constructive feedback, and used digital tools to enhance learning engagement.
Relevant courses:
- Text Mining (COMP61332)
- Agile Software Pipelines (COMP33312)
- Fundamentals of Computation (COMP11212)
- Introduction to Programming 2 (COMP16412)
- Computational Thinking (COMP11012)
- First Year Team Project (COMP10120)

2011 – 2020 Intel Costa Rica · Heredia, Costa Rica

Contributed to multiple corporate STEM programs as a national science fair judge and school workshop instructor; led project proposals that secured two education grants.