Artificial Intelligence Laboratory

DESCRIPTION

TECNALIA's Artificial Intelligence laboratory offers:
- Advanced data/image processing and analysis capabilities.
- Preparation, construction, training, packaging, deployment and monitoring of AI models, as well as the automation of these phases by means of DataOps and MLOps techniques. In addition to the intelligent management of complex analytical pipelines, they can be deployed even on multiple devices in an optimised way.
- Security and reliability of AI models: Anomaly detection, vulnerability assessment, penetration testing, and continuous threat monitoring techniques for the correct implementation of security policies.
- AIOps: Resilience and operationalisation of the software lifecycle using artificial intelligence techniques.
Complex computational problem solving
- Architectures and paradigms for the secure and efficient management and sharing of large volumes of data: Big Data, Data Mesh, Cloud Continuum, Fog/Edge computing, Data Spaces.
- Cybersecurity for AI environments: Cybersecurity assessments and application of AI to improve system cybersecurity and data protection/privacy. Emulation of adversarial tactics to improve the defences of an environment. Actions supported by training and awareness raising activities.
- Automatic adaptation and deployment of AI in embedded edge devices (Embedded Edge AI). As well as automated benchmarking to select the most suitable device for a particular model.
- Data management in impact domains: bioinformatics and sport analytics.
- Deployment of LLMs and augmentation of the developer lifecycle through LLMs: development, testing, debugging, etc.
1) Katea Research Cloud (KRC): It is an open and dynamic platform that allows the creation of a customised cloud infrastructure adaptable to the specific needs of each project. It allows the integration of different platforms, systems and services on it.
2) Katea Data Lab (KDL): A common data platform that integrates information into a cloud-based hub, from which data is governed, accessed and delivered to users, applications and/or other technologies. By consolidating information from different sources, KDL enables a more collaborative and efficient decision-making process for the organisation.
3) Katea Compute Platform (KCP): KATEA's supercomputing centre located in Mikeletegi 2 (Donosti).
4) Katea Data Spaces: experimentation environment to explore and interact with technologies from different data space implementations.
5) 5GEdge Lab composed of a number of devices to allow 5G networks to operate automatically and dedicated embedded devices to explore different artificial intelligence techniques in them promoting low latency scenarios for fast decision making, high privacy and non-connected environments. Devices from three different families microcontrollers, system on chip / system on module and FPGA are available.