MELODIC – Multi-cloud Execution for Data-Intensive Applications

Project acronym:

MELODIC

Programme:

Horizon 2020 – Industrial Leadership – ICT

Funded by:

The European Union

Duration:

1 December 2016 – 31 January 2020

Total project budget:

€4,890,223.75

Funding support for 7bulls.com:

€988,091.25

Website:

https://h2020.melodic.cloud/

CONTEXT AND CHALLENGE

Cloud computing offers scalability and flexibility, but data-intensive applications rarely fit neatly within a single cloud provider. Large-scale analytics workloads often need to balance performance, latency, cost, and data locality across heterogeneous infrastructures.

In practice, cloud ecosystems remain largely isolated. Applications are typically bound to one provider, making it difficult to optimize execution dynamically or to take advantage of favourable conditions across multiple clouds. Manual multi-cloud deployment increases complexity and operational effort, limiting adoption for data-heavy systems.

MELODIC addresses this challenge by enabling data-intensive applications to span multiple cloud environments and to adapt execution decisions automatically as workloads and infrastructure conditions change.

See results and impact

Project objective

The objective of the MELODIC project was to enable large-scale, data-intensive applications to execute efficiently across multiple cloud providers.

The project aimed to automate deployment, optimisation, and orchestration decisions so that applications could select suitable resources dynamically based on performance, latency, availability, and cost, without requiring developers to manage infrastructure complexity manually.

Scope and approach

MELODIC was implemented as a Horizon 2020 research and innovation project running from 2016 to 2020, delivered by a European consortium of research institutions and technology partners.

The project focused on creating a unified execution platform that abstracts heterogeneity across public, private, and hybrid clouds. MELODIC combines model-driven design with optimisation and orchestration mechanisms to support continuous adaptation of applications at runtime.

Within the consortium, 7bulls.com contributed to the design and implementation of core platform components supporting multi-cloud orchestration and optimisation.

Industrial Research

Multi-cloud execution workflow

MELODIC applies an automated workflow for executing data-intensive applications across multiple clouds.

Applications are described using high-level models that capture computational requirements, data dependencies, and quality constraints. Based on these models, the platform evaluates available cloud resources and derives an execution plan optimised for current conditions.

During runtime, monitoring and optimisation mechanisms reassess execution decisions and can trigger reconfiguration or migration between cloud providers, ensuring that applications continue to run in environments that best meet performance and cost objectives.

Technical focus

The MELODIC platform integrates:

model-driven specification of data-intensive applications,
optimisation algorithms for selecting cloud providers and execution plans,
automated orchestration across heterogeneous cloud infrastructures,
intelligent data placement and workload migration mechanisms,
continuous monitoring and adaptation during runtime.

Together, these capabilities enable coordinated, multi-cloud execution of complex data-driven workloads.

Experimental Development

Application domains

MELODIC supports domains where large-scale data processing and dynamic resource optimisation are critical, including:

big data analytics and processing pipelines,
scientific and research computing,
data-driven enterprise and cloud-native applications.

Target users

MELODIC is designed for:

organisations running data-intensive workloads across cloud environments,
teams seeking to reduce vendor lock-in for analytics and processing systems,
architects and operators managing complex, multi-cloud application deployments.

Results and impact

Results and expected outcomes

MELODIC delivered a multi-cloud execution platform enabling data-intensive applications to deploy and optimize themselves across heterogeneous cloud providers based on performance, latency, and cost.

The project shows how model-driven deployment and runtime orchestration can reduce multi-cloud complexity, improve resilience, and support efficient execution of large-scale data-driven workloads. MELODIC also provides the technical foundation for successor projects such as MORPHEMIC.

MELODIC was delivered by a European consortium including:

University of Oslo, CAS Software AG, ICCS, Simula Research Laboratory, University of Ulm, CE-Traffic AS, and 7bulls.com.

Curious about how we can DELIVER in your project?