AI-SPRINT – Artificial Intelligence in a Secure, Privacy-Preserving Computing Continuum

Project acronym:

AI-SPRINT

Programme:

Horizon 2020 – Industrial Leadership – ICT

Co-funded by:

The European Union (H2020 – Industrial Leadership, ICT)

Duration:

1 January 2021 – 31 December 2023

Total project budget:

€4,997,750.00

Funding support for 7bulls.com:

€345,750.00

Website:

https://www.ai-sprint-project.eu

CONTEXT AND CHALLENGE

Artificial intelligence used to be confined to centralised data centres. Now however, AI workloads are increasingly distributed across cloud platforms, edge nodes, on-premises systems, and resource-constrained devices. This shift enables lower latency and new use cases, but it also introduces a critical challenge: how to execute AI securely, efficiently, and reliably across heterogeneous and often untrusted environments.

Traditional AI deployment models struggle to balance performance with strong guarantees around data privacy, security, and isolation. Sensitive data, fragmented infrastructures, and diverse execution environments make it difficult to distribute computation without exposing workloads to risk or losing operational control.

AI-SPRINT was initiated to address this challenge by designing a computing framework that allows AI applications to operate safely across the full computing continuum, without compromising privacy, security, or performance.

See results and impact

Project objective

The objective of the AI-SPRINT project was to develop a secure, privacy-preserving architecture that enables artificial intelligence workloads to be executed across cloud, edge, and on-premises environments.

The project aimed to combine trusted execution technologies, secure data handling, and intelligent workload distribution to support real-time AI processing on heterogeneous infrastructures, including environments that cannot be fully trusted.

Scope and approach

AI-SPRINT was implemented as a Horizon 2020 research and innovation project running from 2021 to 2023, delivered by a European consortium of research institutions and technology partners.

The project focused on creating a programmable computing continuum in which AI applications can be deployed dynamically across distributed infrastructures. AI-SPRINT defines mechanisms that allow workloads to move, adapt, and execute securely wherever computation is most appropriate.

Within the consortium, 7bulls.com contributed to secure cloud-edge integration, distributed execution tooling, and software components that support AI-SPRINT’s programmable continuum.

Industrial Research

Secure execution across the computing continuum

A core element of AI-SPRINT is enabling AI workloads to run safely even in environments that are resource-constrained or not fully trusted. The project introduces mechanisms that isolate computation, protect sensitive data, and ensure controlled execution across distributed systems.

By combining secure enclaves, trusted execution environments, and privacy-preserving data processing, AI-SPRINT makes it possible to execute sensitive AI workloads closer to where data is generated without exposing raw data or compromising system integrity.

‍

Application domains

AI-SPRINT supports secure execution of AI workloads in domains including:

Industrial analytics
Drone monitoring
Personalised services

‍

Experimental Development

Technical focus

AI-SPRINT integrates several complementary technical capabilities to support secure, distributed AI execution:

Privacy-preserving data processing to protect sensitive information during computation
Trusted execution environments and secure enclaves to isolate AI workloads on untrusted infrastructure
Adaptive workload distribution to place computation across cloud, edge, and on-prem systems based on performance and security requirements
Runtime optimisation to support efficient execution on heterogeneous and resource-constrained environments

Together, these components form a cohesive architecture that enables AI applications to operate across fragmented infrastructures as a single, secure computing continuum.

Target users

AI-SPRINT is designed for organisations that need to deploy AI across distributed and heterogeneous environments, including:

enterprises running AI workloads across cloud, edge, and on-premises systems
organisations handling sensitive or regulated data that requires strong privacy and security guarantees
teams developing AI solutions that must operate in real time on resource-constrained or partially untrusted devices

‍

Results and impact

Results and expected outcomes

The AI-SPRINT project delivers a secure and privacy-preserving framework for executing artificial intelligence across the computing continuum. The resulting architecture demonstrates how AI workloads can be distributed dynamically across cloud, edge, and on-premises environments while maintaining strong guarantees around data protection and execution integrity.

Within the project, 7bulls.com’s contribution to secure cloud-edge integration and distributed execution tooling supports practical deployment of AI-SPRINT’s concepts, enabling AI applications to operate reliably across heterogeneous infrastructures under a unified execution model.

AI-SPRINT was delivered by a European consortium including:

Politecnico di Milano
Barcelona Supercomputing Center
TU Dresden
Universitat Politècnica de València
Gregoire SAS
Beck et al Services
Cloud&Heat Technologies
Air Fusion
TRUST-IT
IDC Italia
‍7bulls.com

Curious about how we can DELIVER in your project?

The AI-Sprint project has received funding from the European Union Horizon 2020 research and innovation programme under Grant Agreement No. 101016577