DIGILAB-AI is Ikerlan's applied artificial intelligence laboratory, equipped with advanced infrastructure for AI and Generative AI training, inference, and solution development. The facilities include a training cluster based on state-of-the-art GPUs, a heterogeneous Edge Computing cluster (NPN Preferred Jetson Partner) for distributed processing on embedded devices, and a traditional computing cluster for general support. The laboratory drives innovative projects aimed at digital transformation and industrial competitiveness.
Data
Deployment and Application
Developments
Infrastructure and Computing
AI Training Cluster
DIGILAB-AI's AI training cluster, equipped with NVIDIA’s latest generation GPUs, delivers high performance and efficiency for the fast and scalable training of complex models
Generative AI Training and Fine-Tuning Cluster
DIGILAB-AI's generative AI training and fine-tuning cluster, powered by NVIDIA’s latest generation GPUs, enables the development and optimization of large models with high efficiency and performance.
Heterogeneous Edge Computing cluster
DIGILAB-AI's heterogeneous Edge Computing cluster, developed by Ikerlan as an NPN Preferred Jetson Partner of NVIDIA, integrates embedded devices with advanced processing and inference capabilities, enabling distributed training of AI models in decentralized environments.
Traditional Computing Cluster
DIGILAB-AI's traditional computing cluster includes over 10 hosts configured for virtual machines and Kubernetes environments, providing scalability and flexibility for the development of advanced technological solutions.
Ikerlan offers an AI model inference service leveraging Edge Computing technology and robust infrastructure. The service enables fast and efficient inference in decentralized environments, improving response times and reliability for industrial applications.
Ikerlan provides a customizable service for the development and deployment of MLOps platforms tailored to the specific needs of clients. The service includes automated model training, monitoring, maintenance, and deployment, optimizing the lifecycle of AI in industrial environments.
We provide an evaluation and auditing system specialized in Large Language Models (LLMs), AI agents, and Retrieval-Augmented Generation (RAG) systems, ensuring quality, robustness, and reliable behavior. The service analyzes performance, detects biases, and validates alignment with specific objectives using advanced metrics and customized testing in controlled environments.
Platform for training management of distributed AI models (Vertical Federated Learning and Horizontal Federated Learning). Full lifecycle management of distributed AI models.
Ikerlan offers an AI model training service powered by a robust infrastructure based on advanced GPUs. The service includes the training and fine-tuning of large models, optimizing time and ensuring high performance, tailored to meet industrial and application-specific needs.
We offer a multi-agent platform for the design, integration, and evaluation of intelligent agents that collaborate, reason, and make autonomous decisions, leveraging Generative AI and Large Language Models (LLMs). The service targets use cases requiring content generation, advanced coordination, and adaptability, integrating cutting-edge technologies in Machine Learning, Intelligent Planning, and Natural Language Processing (NLP).
Ikerlan provides companies with a comprehensive R&D project service to drive the development and industrial transfer of innovative technologies. The service includes needs definition and analysis, the design of customized technological solutions, prototype development and testing, and support for market implementation, tailored to the specific demands of each sector.