An innovative language model that interprets your data securely.
PrivateAI Suite
The project
PrivateAI Suite is a language model designed for businesses that wish to use artificial intelligence without compromising the confidentiality of their data.
We have designed and developed the platform’s interface: a simple and intuitive space from which to interact with a private AI, trained on the information and requirements of each client company.
The solution is suitable for a range of sectors and enables texts to be interpreted and generated accurately, reliably and responsibly.
The challenge
More and more companies want to integrate AI into their processes, but many cannot afford to expose sensitive data to external services.
We needed an interface that would make a private AI model accessible and straightforward to use, even for those without technical expertise, whilst maintaining high security standards and allowing it to evolve over time in line with the client’s needs.
Our process
We worked on two fronts: on the one hand, a design system that would make the user experience clear and consistent; on the other, an application capable of interacting seamlessly with a bespoke AI system tailored to each company.
We have created a design system that is aligned with the client’s brand identity and designed for companies wishing to use a private and secure AI model. The interface is minimalist, simple and intuitive, ensuring a seamless user experience even for those new to AI.
We have developed the PrivateAI Suite as a PWA (Progressive Web Application), using React and Next.js as the main library and framework. The application’s functionality is based on REST API calls, which enable communication with an AI system that is trained on a case-by-case basis using the client company’s information and requirements. This approach allows for continuous customisation and seamless interaction.
Interface design built in
2 Months
Codebase for every device
1
Our approach
Our scope was clear: the user experience. The model and the data remain in the client’s hands; we focused on everything that lies between the AI and the people who use it.
We took the existing identity and translated it into a system of components, so that every new feature could be developed in a way that was already consistent with the rest.
We designed the product with the idea that it would evolve over time. Every company trains the model in a different way, and the interface needed to adapt without having to be redesigned each time.
Technologies
- Next.js
- React
- REST API
