QUANT AI Lab

Madrid, Spain

Contact for ratesResponds in 24h

About

QUANT AI Lab is a Madrid AI company building risk, forecasting and decision systems for regulated and operational environments, with offices in Paris, Bordeaux and London.

Our Take

Twenty-eight published use cases, and what makes them worth reading is that each one names its stack.

Fraud detection on Databricks, Neo4j and Snowflake. A destination recommender for airline revenue built with SageMaker, Hugging Face, TensorFlow and Airflow. Retrieval-augmented chat on LangChain and FAISS. A Bayesian library industrialized with pgmpy, Docker and FastAPI. Climate transition risk, ESG flood-risk assessment against CSRD, PPE detection, predictive maintenance, unsupervised fraud detection. That is an engineering catalog rather than a marketing one, and it is the most technically itemized portfolio on this register.

The practice is anchored in quantitative finance—the group chief executive is a known figure in model risk—and Virgo, its own architecture, is aimed at regulated and mission-critical deployment.

The weakness is attribution. BNP Paribas, Santander, BBVA and Repsol appear as logos, every use case is anonymized, and none carries a result. The technical credibility is high and the delivery record is unverifiable from outside. Ask which logo maps to which build.

Key strengths

Twenty-eight use cases each naming its technology stack, which lets an engineering buyer assess the team before the first call
Genuine quantitative depth—Bayesian networks, hierarchical models, optimization—rather than an LLM-only practice
A registered Spanish company with a named executive team and offices in Madrid, Paris, Bordeaux and London

Good to know

Every use case is anonymized and the four bank logos attach to none of them—ask which client corresponds to which build
No outcome figures anywhere; the catalog proves capability and says nothing about impact