The recent power supply interruption in Spain has highlighted the urgent need to improve predictive capacity and strategic decision-making in energy systems. In this context, the technology developed by Spanish startup AyGLOO positions itself as an essential ally for network operators and energy managers.
Who is AyGLOO?
AyGLOO is a Spanish technology company specialized in Prescriptive Decision AI, an emerging discipline that makes AI model results transparent and understandable. Their platform integrates with existing customer systems without replacing them and provides traceability, detailed explanations, and advanced analysis tools that allow technical and business teams to understand and audit each prediction. The development of this AyGLOO technology has financial support from the Center for Industrial Technological Development (CDTI).
Prescriptive Decision AI to anticipate critical events
AyGLOO's claims are supported by numerous scientific studies, including a recent one that demonstrates how Prescriptive Decision AI technology improves wind energy prediction by providing a detailed understanding of prediction models, a study developed by the Wind Engineering and Renewable Energy Laboratory (EPFL_WIRE) of the Swiss Federal Institute of Technology Lausanne (EPFL).
The use of this technology is highly relevant for reducing Spain's and the EU's energy dependence, optimizing the integration of renewable sources into the system.
The AyGLOO platform offers unique functionalities:
• Critical segment detection where the model fails or generates uncertainty. • Simplified twin models that simulate alternative scenarios, including exogenous factors, without the need for retraining • Multi-scale explainability, offering analysis from a global vision to hourly detail. • Impartiality and robustness analysis, including bias detection and data misalignments. • "What-if" and counterfactual scenario simulation, to analyze how variations in factors (such as temperature or electricity price) affect predictions. • Complete decision traceability, ensuring that each prediction is understandable and auditable.
AyGLOO has developed technical innovations such as dynamic surrogate models and intelligent segmentation algorithms capable of extracting powerful insights that can prevent extreme events like the one recently experienced in the Iberian Peninsula.
Transform data into decision
Unlike approaches that prioritize only accuracy through "black box" models, AyGLOO provides the intelligence layer that transforms predictions into useful, explainable, and actionable knowledge. Their proposal allows operators to anticipate problems, adjust their existing models, and make informed decisions even in conditions of high uncertainty, such as those experienced during the recent blackout.
As AyGLOO's CEO, Ignacio Gutiérrez Peña, emphasizes: "Our technology transforms predictive models into strategic tools, allowing electricity system managers to confidently anticipate critical events like the recent blackout. It's not just about predicting, but understanding why, how, and when something will happen."
In short, AyGLOO doesn't compete with traditional machine learning models, but rather enhances them. Facing an increasingly complex and volatile energy environment, solutions like theirs allow moving from an isolated prediction to a complete, auditable, and actionable story that strengthens the resilience of the electrical system.
In addition to their focus on the energy sector, AyGLOO leads other innovative R&D projects, such as one with Europa Press and the Autonomous University of Madrid to develop a tool that detects fake news and explains why they are fake, applying AI and Prescriptive Decision AI techniques.