Advanced the Expert Brain initiative in collaboration with VRAIN, focusing on multi-agent systems, LLM integration, and industrial process optimization for additive manufacturing.
Built a SPADE-based multi-agent deep research system that autonomously retrieved sources, synthesized findings, and generated structured cited reports.
Maintained SPADE-LLM while leading its package migration to uv and CI overhaul, introducing features like structured outputs, RAG/MCP, and saving ~6,000–9,000 GitHub Actions minutes annually.
Co-developed a Bayesian optimization tool for LPBF printing that recommended process parameters and proposed next-best experiments to maximize part density.
Selected as one of 90 participants from more than 105,000 applicants for the inaugural Santander Open Academy programme, representing more than 13 countries.
Our multidisciplinary team won the Final Pitch Challenge after pivoting and rebuilding our proposal under significant time pressure.
Built a strong mathematical foundation during this year before moving into Data Science to focus more directly on applied machine learning, AI, and independent projects.