Frugality applied to the digital sector, particularly in the context of artificial intelligence, marks a profound shift in how we approach technology design. As discussed in this article by Tina Issa, it’s no longer just about pursuing raw performance or algorithmic sophistication, but about making more responsible technological choices, better aligned with current environmental, economic, and social constraints.
However, frugality cannot remain a theoretical principle or laudable intention. To have a real impact, it must be operationalised and translated into concrete methods, tools, and practices. This is precisely the ambition of the E3S research axis – Efficient & Sustainable Smart Solutions, led by Devoteam Research & Innovation.
This axis aims to structure and develop intelligent solutions capable of helping organisations and individuals make the right digital choices, reconciling efficiency, value creation, and sustainability throughout the entire lifecycle of systems.
Why Efficient & Sustainable Smart Solutions Today?
The digital sector now represents a significant share of global energy consumption and carbon footprint. At the same time, companies face sometimes contradictory demands:
- Innovate faster
- Exploit ever more data
- Integrate AI at scale
- While reducing costs and environmental impact
Faced with this tension, two pitfalls are common:
- The “all AI” approach, where artificial intelligence is used even when simpler approaches would suffice
- Untooled frugality, reduced to best practice guides that are difficult to apply in daily operations
The E3S axis starts from a clear observation: to make digital technology truly more sustainable, we must equip decision-making, support usage practices, and put intelligence in service of digital sobriety, not the other way around.
E3S is built on two fundamental pillars:
- Efficiency, understood as the capacity to maximise created value while minimising effort, cost, and mobilised resources
- Sustainability, which integrates environmental, social, and economic dimensions into the design and use of digital solutions
The strength of the E3S axis lies in the synergy between these two dimensions, avoiding treating them separately.
A Global Vision, Implemented Through Complementary Projects
To address these challenges, the E3S axis has been structured around four research projects, each acting at a different level of environmental and organisational impact leverage in digital technology.
These projects are not independent silos. On the contrary, they form a coherent ecosystem, covering:
- Value creation through data
- Education and dissemination of best practices
- Concrete software development practices
- And technological choices regarding frugal AI
AI4Value: Creating Value Efficiently from Data
Many companies today possess large volumes of data, often heterogeneous and underutilised. Yet, rapidly identifying relevant AI use cases that truly create value remains a complex, lengthy, and largely manual exercise.
The AI4Value project tackles this problem by proposing a radically different approach: starting from the data itself, rather than a predefined use case, to automatically generate AI use case recommendations with high potential.
The objective is to develop an intelligent assistant capable of:
- Automatically analysing raw and varied datasets
- Extracting actionable insights
- And proposing contextualised, relevant, and value-generating AI use cases
AI4Value fully aligns with the E3S logic by seeking to reduce analytical effort waste, improve data team efficiency, and direct AI investments toward truly useful and measurable use cases.
Green4ALL: Democratizing Green IT and Sustainable IT Best Practices
While Green IT best practice frameworks exist, their adoption often remains limited. They are perceived as complex, poorly contextualised, and mainly intended for experts.
The Green4ALL project aims to remove this barrier by developing a comprehensive intelligent assistant capable of making Green IT and Sustainable IT best practices:
- Accessible to the widest audience
- Contextualised according to role, expertise level, and user context
- Directly actionable in daily professional activities
Green4ALL addresses both technical and non-technical profiles. It serves as an educational and action tool, supporting users in their digital decision-making and practices.
This project plays a key role in the E3S axis: it enables the widespread dissemination of digital sobriety culture, an essential condition for sustainably transforming practices.
GreenSoftware: Acting at the Heart of Code for More Sober Digital Technology
Every line of code has an impact. Unnecessary CPU cycles, excessive memory consumption, or redundant processing contribute significantly to a substantial environmental footprint at scale.
The GreenSoftware project directly addresses this technical lever by developing an intelligent assistant capable of:
- Analysing existing code
- And automatically proposing more sober code alternatives while maintaining functional performance
The ambition is twofold:
- Reduce cognitive effort for developers by integrating sobriety directly into their workflows
- Promote the mass adoption of sustainable development best practices, which are often undervalued or difficult to apply manually
GreenSoftware embodies the E3S approach at the most operational level: making sobriety a development reflex, supported by intelligent and measurable tools.
SustAInable: Recommending Truly Frugal AI Solutions
With the proliferation of AI models and architectures, a major challenge emerges: How to choose the right solution, at the right time, with the right level of complexity and impact?
The SustAInable project addresses this issue by developing a frugal AI solution recommendation assistant. Its objective is to guide users toward:
- AI or non-AI building blocks adapted to their real needs
- Taking into account technical, economic, and environmental constraints
- And adjusting the recommendation level according to user expertise
SustAInable follows a logic of informed decision-making, avoiding systematic recourse to heavy models when simpler, more sober, and equally effective approaches are possible.
This project embodies the E3S philosophy: utilising AI to support more responsible choices, rather than contributing to uncontrolled technological expansion.
Overall Coherence in Service of Responsible Digital Technology
Taken individually, each of these projects addresses a specific challenge. Together, they form a complete value chain for responsible digital technology:
- AI4Value acts on efficiency and value creation through data
- Green4ALL disseminates best practices and facilitates their adoption
- GreenSoftware transforms development practices
- SustAInable guides technological choices toward frugal solutions
This complementarity is at the heart of the E3S axis. It enables simultaneous action on decisions, practices, and implementations, with a common objective: building high-performing, sustainable, and intelligent digital systems.
Efficient & Sustainable Smart Solutions, a Structuring Framework for Tomorrow’s Digital Technology
The Efficient & Sustainable Smart Solutions axis illustrates a strong conviction of Devoteam Research & Innovation: the transition to responsible digital technology will not happen through constraint alone or technology alone, but through an intelligent combination of both.
By equipping frugality, supporting practices, and integrating intelligence at the heart of decisions, E3S proposes a pragmatic and actionable path toward more sustainable digital solutions.
In upcoming articles, we will revisit each of the axis projects (AI4Value, Green4ALL, GreenSoftware, and SustAInable) in more detail to explore their scientific challenges, technical choices, and impact perspectives.
Because the future of digital landscape is not solely about power, but about our collective capacity to make the right choices.