About Luis A.
I help researchers, startups, and technical teams turn math-heavy ideas into clean, reliable ML systems—from research prototypes to production-grade LLM/RAG applications. My work sits at the intersection of probabilistic ML, deep learning, and robust engineering, with a strong focus on correctness, evaluation, and maintainability.
Strengths: bridging theory ↔ implementation; designing uncertainty-aware methods; implementing in Python/PyTorch; delivering well-tested modules with clear docs and reproducible results.
LLM/RAG delivery:
- End-to-end RAG pipelines: ingestion → chunking → embeddings → vector search → re-ranking → citation/attribution → evaluation
- Guardrails, monitoring, security-minded design, and cost/latency optimization
- Integration with existing stacks: APIs, databases, auth, and data governance
Highlights:
- PhD in AI/ML.
- Peer-reviewed publications (ICLR/ICML/AISTATS/NeurIPS/JAIR)
- Open-source PyTorch implementations adopted by multiple academic groups
- University teaching (Python programming, Bayesian methods) and high-quality technical documentation
Team option (AI + Architecture)
For full-stack AI delivery I partner with Jesús Segado Fernández (#jesussegadofernandez) (Senior Software/DevOps Architect): — he covers platform architecture (Kubernetes, CI/CD, IaC, security, observability) while I lead the ML/LLM layer.
English
Native or bilingual
Spanish
Native or bilingual
Remote only
Primarily works remotely
Experience
- Aalborg UniversityMachine Learning ResearcherRESEARCHFebruary 2026 - Today (5 months)Copenhagen, DenmarkDevelop and apply probabilistic machine learning methods for large-scale biological and environmental data. The work includes uncertainty-aware representation learning, clustering, and prediction, with applications in metagenomic genome recovery, microbial biogeography, and ecological network analysis. The role combines methodological research in Bayesian inference and deep learning with close collaboration across genomics, ecology, and data science.
- ToptalMachine Learning ConsultantDIGITAL AND ITDecember 2025 - Today (7 months)Provides end-to-end consulting for LLM-based systems, covering RAG architecture, model selection and fine-tuning, vector search, prompt and retrieval optimization, and robust MLOps practices. Emphasis is placed on monitoring, evaluation, security, and operational stability in enterprise environments.
- Autonomous University of Madrid,Research Scientist (PhD)December 2021 - December 2025 (4 years)Spain
- Conduct research on Bayesian deep learning, uncertainty quantification, and PAC-Bayes general ization bounds for modern deep networks.
- Developed and released Variational Linearized Laplace and post-hoc uncertainty estimation meth ods in PyTorch, adopted by 5+ academic research groups.
- Collaborated with the University of Cambridge (LLM uncertainty quantification) and Aalborg University (PAC-Bayes/Chernoff bounds).
- Published 3 first-author papers at ICLR, ICML and AISTATS; and 2 second-author papers at NeurIPS and JAIR (with oral talk at ECAI 2025).
- Taught Python Programming (BSc) and Bayesian Methods (MSc) courses.
- Tracked experiments with MLFlow and deployed Docker images using Google Cloud & AWS.
Recommendations
Be the first to recommend Luis A.
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Ph.D. in Machine LearningAutonomous University of Madrid2025Ph.D. in Machine Learning
- M.S. Data ScienceAutonomous University of Madrid2022M.S. Data Science
Certifications
- Building RAG Agents with LLMs – Developed Retrieval-Augmented Generation (RAG) agentsNVIDIA2025