Open to opportunities

Manu FIC-electronica

@manufic-electronica

Message

AI researcher and software engineer building NLP methods for early mental-health risk detection from social media.

Spain

Message

What I'm looking for

I’m looking for an AI/NLP role where I can build NLP for early mental-health risk detection, ship reproducible research pipelines, and collaborate with teams that care about real-world impact.

I’m an AI Researcher & Software Engineer with a PhD in AI (Cum Laude), focused on NLP for early risk detection on social media. I develop computational methods to detect depression, anorexia, self-harm, and gambling risks across 18 eRisk (CLEF) datasets.

My work includes a temporal word-embedding framework that tracks semantic drift in users’ posting histories, achieving F1 0.84 on gambling detection (eRisk 2023). I also evaluate topic models with human judges—BERTopic significantly outperformed LDA and TopClus—and relate discovered topics to depression indicators via BDI-II analysis.

I care deeply about reproducibility and practical impact. I built and open-sourced LabChain (SoftwareX), a modular pipeline with deterministic hash-based caching that surfaced a hidden preprocessing bug, delivered performance gains up to 192%, and eliminated ~48h of redundant computation in production research workflows at CiTIUS.

Experience

Work history, roles, and key accomplishments

Current

PhD Researcher (NLP)

Current

CiTIUS, Universidade de Santiago de Compostela

Mar 2022 - Present (4 years 4 months)

Developed and evaluated computational methods for early detection of mental-health conditions from social media text across 18 eRisk (CLEF) datasets. Proposed temporal word-embedding approaches and evaluated topic modeling methods to identify depression indicators.

Python PyTorch Transformers NLP Rust PyO3 Axum SvelteKit TypeScript Tauri fastAPI Next.js Docker Linux Git Temporal Word Embeddings Topic modeling

Research Support Technician

CITIC, Universidade da Coruña

Oct 2021 - Apr 2022 (6 months)

Applied machine learning research for NLP (PyTorch) and supported work prior to starting a PhD. Worked on a pipeline framework with hash-based caching for reproducible ML experiments.

Python PyTorch NLP Hash Based Caching Reproducible ML

Research Assistant

Universidade da Coruña

Apr 2021 - Oct 2021 (6 months)

Applied machine learning research (PyTorch) in the university’s IR Lab. Built a distributed runtime for research pipelines and agent orchestration, including Bayesian/PBT studies and an MCP server.

PyTorch Machine Learning Research Distributed Computing Agent orchestration MCP Servers Rust PyO3

Junior Software Developer

Grupo Oesía

Apr 2017 - Apr 2018 (1 year)

Progressed from intern to junior developer and contributed to enterprise software development using Git and Linux. Built, compared, and evaluated LLM assistants, including a visual Agent Studio with parallel graph execution.

Git Linux LLM Parallel Graph Execution

Education

Degrees, certifications, and relevant coursework

Universidade de Santiago de Compostela

PhD, Information Technology Research

2022 - 2026

Grade: Sobresaliente Cum Laude

PhD in Information Technology Research (Sobresaliente Cum Laude). Thesis: Online Content Mining to Improve Psychological Profiling of Individuals under supervisors David E. Losada (USC) and Javier Parapar (UDC).