Targa Telematics accelerates AI adoption with an on-premise software development environment based on open-weight models

Targa Telematics, one of the world’s leading players in AIoT (Artificial Intelligence of Things) for the development of digital solutions and platforms for connected mobility, announces that it has completed and deployed an on-premise software development environment based on generative artificial intelligence (LLMs) and open-weight models. The solution is designed to support its development teams and accelerate innovation across the company.

Delivered on a Hewlett Packard Enterprise (HPE) platform in collaboration with Gruppo Zenit, a Managed Service Provider specialising in the design and management of high-performance data centre and networking infrastructures, the project has resulted in an enterprise AI platform used by around 150 engineers. The objective was to create a development environment that combines high performance, full control over data, security and the flexibility to evolve further.

The platform leverages open-weight AI models, enabling inference to be performed entirely within the company’s own infrastructure, with no need to transfer data or proprietary information to public clouds or third-party services. This allows Targa Telematics to benefit from the pace of innovation offered by the open-source ecosystem, while maintaining control over its information assets, model governance and the customisation of models to meet specific business requirements.

As companies are balancing the rising adoption of generative or agentic AI with the need to protect their information assets, Targa Telematics has developed a fully on-premise platform, so that no company data is transferred to external infrastructures or public clouds.

Primarily used to support coding activities, the platform also acts as an AI assistant. By running LLMs within the company’s own environment, it ensures that information always remains under Targa Telematics’ control. This approach is instrumental in addressing growing regulatory and market requirements around data sovereignty, cybersecurity and compliance, while maintaining the flexibility needed to experiment with new AI models.

Targa Telematics’ initiative was driven by a clear objective. “Generative AI is increasingly becoming a day-to-day tool for our developers. We wanted to create a fully on-premise development environment while retaining complete control over our code, data and technological know-how,” said Carlo Stefanelli, CTO of Targa Telematics. “An on-premise approach, based on open-weight models, is key to ensuring that all information remains entirely within our corporate environment, allowing us to foster innovation without compromising on data security and control.”

Gruppo Zenit contributed to the design of a robust infrastructure focused on security, business continuity and long-term data governance. The solution was built using Hewlett Packard Enterprise (HPE) technologies, which provide the computing power required for AI inference alongside the tools needed to guarantee that the on-premise environment is reliable, scalable and securely managed.

“AI can only generate value when the underlying infrastructure is designed to be secure, scalable and manageable. Our role is to support companies in developing ecosystems where innovation and data protection can grow and flourish hand in hand. The project we have carried out with Targa Telematics shows how AI experimentation can be transformed into a platform primed to drive business evolution,” said Giuseppe Barbera, CEO of Gruppo Zenit.

The results demonstrate the value of Targa Telematics’ approach. In the first seven weeks after deployment, the platform handled almost 30,000 requests from developers, processing more than 3.4 billion tokens and generating 65 million tokens in response. These figures highlight the robustness of the on-premise infrastructure, as well as the rapid adoption of the solution by the company’s software development teams.

Through this project, Targa Telematics is strengthening its expertise in generative AI, developing an on-premise infrastructure model that can be replicated for future use cases.