The goal of the DHIRIA project is the design and development of a Cloud-based platform to provide machine and deep learning “as-a-service” solutions able to process encrypted data by means of Homomorphic Encryption mechanisms. In more detail, the ability to provide advanced machine and deep learning solutions in an "as-a-service" mode and the ability to process user data in an encrypted manner are the two distinctive and highly innovative features of the DHIRIA project.
These “as-a-service” solutions provided by the DHIRIA platform will be made available in three different configurations:

1) PaaS - Platform-as-a-Service: models are trained by users and made available in a B2B and B2C manner through the platform and Cloud services offered by DHIRIA.

2) SaaS - Software-as-a-service: ready-to-use solutions are made available as a product/service to users

3) PP-SaaS - Privacy Preserving Software-as-a-Service: ready-to-use solutions solutions are made available as a product / service operating on encrypted data.

Activity:
Design and development of artificial intelligence operating on encrypted and plain data in platform-as-a-service and software-as-a-service mode.

Market:
The reference market in which DHIRIA will operate comprises more consolidated but still expanding market areas (AI and Cloud) and emerging areas such as Analytics- as-a-service and Homomorphic Encryption (HE) with annual growth rates ranging from +9.7% to 44.3%, while the potential target scenarios comprise finance, energy, IT, content creation and health. The competitive scenario comprises big tech companies characterized by a “limit pricing” policy and providing mass-market services, hence leaving room for novel and cutting-edge solutions (in particular for SMEs). This is particularly true for the privacy-preserving machine and deep learning computation representing a completely new market with only very few competitors in US (mainly start-ups) leaving huge opportunities of growth.

Contacts:
manuel.roveri@polimi.it
info@dhiria.com