Success Cases

Logistics & Transport

Andalusian Logistics Group

Challenge: Optimization of distribution routes with traffic constraints, time windows and fuel cost.

Solution: We developed a route prediction system based on reinforcement learning (RL) integrating historical traffic data, weather conditions and demand patterns.

Results:

• 22% reduction in fuel consumption.
• 18% improvement in delivery punctuality.
• ROI achieved in 7 months.

Healthcare

Regional Hospital

Challenge: Overload on initial triage phone lines, long wait times and high administrative burden.

Solution: Implementation of a conversational chatbot with fine‑tuned language model for frequent medical questions, integrated with the electronic health record system.

Results:

• 35% reduction in call center calls.
• 25% improvement in patient satisfaction (surveys).
• 92% correct redirection of cases to appropriate specialists.

Financial Services

Credit Institution

Challenge: Detection of fraudulent transactions in real time without increasing false positives.

Solution: Deployment of an anomaly detection model based on variational autoencoders (VAE) analyzing transactional behavior patterns.

Results:

• 40% more fraud detection than previous system.
• 60% reduction in false positives.
• Real‑time processing (<100 ms per transaction).

Retail

Supermarket Chain

Challenge: Inventory optimization and demand forecasting to reduce waste and stock‑outs.

Solution: Multi‑scale forecast model (daily, weekly, monthly) with recurrent neural networks (LSTM) and sales, promotion and local events data.

Results:

• 30% reduction in perishable goods waste.
• 12% increase in sales due to optimized availability.
• 80% automation of supplier orders.