Engineering announces the new version of EngGPT, its proprietary Large Language Model (LLM) designed to meet the needs of businesses and public administrations with even greater precision. This latest version ensures a secure and AI Act-compliant Private GenAI model tailored to specific industry requirements.

"EngGPT and its new version represent a key milestone in our commitment to supporting Italy’s innovation ecosystem and digital transformation. We look to the future with the firm belief that Italy can continue to be a leader in AI by fully leveraging its potential," says Maximo Ibarra, CEO of Engineering, a leading company in the digitalization of processes for enterprises and public administration. "Our approach seamlessly combines research and business, integrating technological and multidisciplinary expertise. This enables us to design a cutting-edge Generative AI solution that meets the specific needs of the major players we collaborate with, generating tangible and sustainable value for the future of all."

Already adopted by numerous organizations undergoing digital transformation with Engineering's support, EngGPT now evolves with a new architecture/model that further enhances response accuracy. The model captures linguistic nuances, idiomatic expressions, and the diverse economic, social, industrial, and regulatory contexts of Italy. This evolution allows Engineering to provide both public and private entities with a GenAI solution tailored to specific sectors and applications.

The project is part of the ICSC - National Research Center for High Performance Computing, Big Data, and Quantum Computing, funded by PNRR (Italy's National Recovery and Resilience Plan), where Engineering is one of the founding members. It also benefits from the computational power of CINECA, a recognized Italian and European center of excellence in supercomputing technologies for frontier applications.

Thanks to CINECA’s computational capacity, the new version of EngGPT is trained "from scratch" with trillions of tokens optimized for the Italian language. This approach ensures full control over the training data, minimizing biases and guaranteeing a highly accurate and secure model that adheres to the highest standards of quality and transparency. The model is increasingly aligned with AI Act guidelines, ensuring an AI system that safeguards citizens' rights.

The evolution of EngGPT, whose training phase is already underway, is part of the broader strategy of Engineering, led by Maximo Ibarra. With over €30 million invested annually in research and innovation, the company continues to develop advanced solutions for Italy’s digital transformation. EngGPT represents the intersection of research, expertise, and business, bringing together a multidisciplinary team of around 70 professionals, including data scientists, legal experts, physicists, humanities scholars, neurolinguists, and lawyers. This collaborative approach ensures that technology, given its profound impact on society, is managed in an ecosystemic, ethical and interdisciplinary manner.

Engineering has been working in Artificial Intelligence since 1987, when it launched its first AI laboratory. Today, its AI Competence Center includes more than 400 professionals, supporting over 150 clients across businesses and public administrations in adopting and fully utilizing AI, GenAI, and Data Analytics applications. Additionally, the company has a dedicated research team that, between 2020 and 2025, has developed over 110 Research & Innovation projects, integrating AI and Advanced Analytics technologies. About 40 of these projects have a strong focus on AI applications in strategic sectors, such as Energy, Manufacturing, Public Administration, and Healthcare, with a total investment in the last five years exceeding €32 million.