Carla Atienza Vilela

I’m an AI graduate from UPC Barcelona (October 2026), focused on machine learning, NLP and language models, from data preparation through training, evaluation and interpretation.

At Hiberus I worked on an AI travel-itinerary project for El Corte Inglés. I built a semantic-matching system with embeddings, tested the platform end to end to trace errors to their source (front-end, back-end, data or model), designed consistency tests to reduce LLM hallucinations, and wrote the user manual. I developed in C# (.NET).

My thesis (graded 9.5/10) detects Alzheimer’s from speech on the Pitt DementiaBank corpus. I compared a classical ML model with linguistic features, a few-shot LLM and a hybrid of both (AUC 0.87), and analysed where and why each one fails. Other projects include a multi-agent medical diagnosis system with Telefónica (LangGraph, LoRA fine-tuning), a pulmonary fibrosis prediction model, and NER systems in Spanish and Dutch.

I have solid SQL, NoSQL and MongoDB foundations, plus experience cleaning and unifying datasets. I’m a native Catalan and Spanish speaker (C1 English and French), and I spent a year at Politecnico di Torino on Erasmus taking Data Science master’s courses.

I’m looking for remote roles in AI engineering, data analysis, QA for AI products and LLM evaluation, ideally where I can keep learning and work on real problems.