The digital business ecosystem GAIA-X – A session of the FINDER inclusive digital innovation week

At the second day of our Inclusive Digital Innovation in Financial Services & Insurance (FSI) event, we had a look on a European moon-shot project GAIA-X. For this topic we were happy to welcome Hubert Tardieu, chairman of the Board of GAIA-X, highlighting how the project will shape the future of the financial services and insurance market in Europe by “creating a next generation data ecosystem for Europe with a global aspiration”.

The kick-off summit of GAIA-X in 2020 consequentially focused on two major foundations for a project of such scale. First, the overall key concepts to achieve the envisioned cloud penetration of and in the European market were presented. These depict the five pillars of GAIA-X:

  1. Supporting policy rules derived from requirements of a European single market
  2. Support federal data infrastructures (methodology to synthesize different frameworks)
  3. Ensuring interoperability, sovereignty, portability of data
  4. Providing testable compliance to GAIA-X Architecture of standards
  5. Acknowledging open standard setting processes laid out in the internal GAI-X rules of the GAIA-X

Second, the projects governance structure was outlined. Both points reveal what is at the heart of GAIA-X: creating a digital business ecosystem for open innovation.

GAIA-X – digital business ecosystem by design

The professional literature[1][2][3] highlights a couple of design principles to achieve a successful setup of a healthy digital business ecosystem. The summit therefore was existential to growing legitimacy by presenting that GAIA-X is shaping its governance based on these, interdependent, design principles:

  • Demand orientation; stating a mission allows enthusiastic actors to push into the ecosystem instead of pulling them in securing pro-active, responsive behaviour for the joint value creation
  • Openness; in form of a transparent environment enabling an easy access
  • Self-organization; enabling participants to act autonomously to increase commitment
  • Loose coupling; so that participants can join freely and engage in open relationships so that there are no heavy dependencies determining the success of conducted projects
  • Domain clustering; enabling the grouping of participants in projects based on shared interests

Advantages of GAIA-X as a digital business ecosystems

Mr. Tardieu sees a major benefit of GAIA-X in the enabled Europe-wide collaboration between private and public sector. First, the initiative allows to jumpstart the facilitation of digital competence of European companies and thus of Europe as well. Second, it enables European-centred research hand-in-hand with practitioners which in return enables more precise policy-making. Third, the collaboration allows for a holistic approach since it is the only project that is addressing all the necessary elements together from root-to-tip. The initiative captures the alignment of technical standards and services for interoperability and portability (the roots) through the federated trust and sovereignty services (trunk) up to the definition of ontologies, APIs and technology standards for compliance (leaves).

“Gaia-X may seem gigantic but we don’t think that the big issues we are trying to tackle can be ‘sliced’ into smaller parts. In Financial Services and Insurance, this is especially true at a time when cloud adoption needs to accelerate and there is so much change within the industry.”

Based on these the concept of open strategy, nested in a digital business ecosystem, offers advantages for both affiliated producers and consumers.[4] Since for the GAIA-X participants are often both – provider and user of data – the approach taken by GAIA-X is of particularly high functionality. Benefits for data providers are lowered development and launch costs, quality improvement due to a joined development environment and increased speed to market. This, in turn, translates into the benefits of data users since the reduced costs are reflected in the price (up to being open source) as well as a direct incorporation of feedback and implementation of specifications in the development cycle.[5]

Challenges identified, faced and tackled

Nevertheless, open strategizing also presents those involved with the challenge of finding of losing established business models of value appropriation. In the case of GAIA-X, there are two factors that endanger common business models of participants:

  1. The lower costs (should costs be charged) of developed services or products are passed on directly to the user of the data. This significantly reduces the achieved profit.
  2. Differing ownership of input data, managed through data sharing agreements and data use statements, impedes the distribution of the benefits achieved.

Hence, participants have to find new ways to appropriate value within the value chain and thus generate profit from their engagement in the digital business ecosystem.[6][7] According to Mr. Tardieu the participants at GAIA-X are fully aware of that challenge:

Of course, our concern is how do we create data spaces in which those involved will be able to further their own interests too. We also think about how those who put the most effort in from the start don’t lose out to those who might join later when the hard work is done. We reject the idea of selling data. It is an old-fashioned way of thinking.

Part of this process is the integration of researchers as they are developing novel ideas to tackle this challenge. A recently idea explored is the approach of ‘Tickenomics’, originating at the University of Toulouse. The underlying mechanism is illustrated by Mr. Tardieu through an analogy:One example [of Tickenomics] would be to suppose you are in a place with no transportation system. You are selling tickets (or lots) to travellers, to towns and to whole regions. At some point, you will have enough money to create the transportation system. And that is when the tickets become valuable. It might be slow getting started but as soon as everything is in place, it takes off quickly. So we are looking at ways we might introduce ‘tickets’ without the possibility of these leading to monopolies.”

Furthermore, open strategizing[8] allowed the participants of GAIA-X to identify the following barriers which hinder the successful realisation of the visionary mission:

  1. The absence of portability (also known as vendor lock-in or the risk of ‘mainframe syndrome’) – preventing companies from committing to cloud due to future risk.
  2. The potential lack of interoperability – whereby differences in the technical infrastructure may hamper or even render data sharing impossible
  3. The importance of data sovereignty – as otherwise companies would refrain from moving to the cloud due to the risk of misappropriation of shared data.

At the same time the mission-driven initiative also produced a mode of operation to tackle these barriers according to Mr. Tardieu: “The challenges of data portability, interoperability, and common commercial and legal frameworks have different implications for different industries. That is where GAIA-X is helping participants come together to define use cases for their industries and to share information that can make industry data spaces possible. This collaboration is important.”

GAIA-X enables mission-oriented work in industry-specific projects

With the mission, design and barriers of the digital business ecosystem being fleshed out, naturally, the question occurs how GAIA-X will manifest itself through the realisation of projects. Mr. Tardieu pointed out that therefore the domain clustering is of importance as it defines groups to create “data spaces” on an industrial level. Therefore, the FSI industry is a prime pilot since, due to the high level of regulation, collaboration between participants of the public and private sector is required when tackling the challenges of data portability, interoperability, and common commercial as well as legal frameworks.

By working together, they (public & private actors) can increase the chance of success. And this isn’t just about sharing data, remember. It’s also about infrastructure too. Especially where regulations dictate compliance at a local level. You can’t just do it at the application level. This is something GAIA-X is working on.

Furthermore, he lined out that the FSI industry “is ‘ahead of the pack’ because of PSD2 (for a brief overview of PSD2 see this blogpost). We wouldn’t have seen the huge development of FinTechs without it. But this is only half of the work. Data ontologies are key and you will soon see the priority use cases from the financial services and insurance sector start to emerge based on GAIA-X projects.”

The first pilot project – the safe Financial Big Data Cluster – investigating the use case of a joint platform to fuel artificial intelligence services, is currently developed by participants of the private and public sector with involvement of FINDER (for an introduction see this blogpost).

Added benefit to the FSI industry through GAIA-X

While there are different initiatives (for instance, the EU Alliance for Industrial Data and Cloud) Mr. Tardieu sees the benefit of GAIA-X in its holistic approach since it is the only initiative that is addressing all the necessary elements together from root-to-tip. The initiative captures the alignment of technical standards and services for interoperability and portability (the roots) through the federated trust and sovereignty services (trunk) up to the definition of ontologies, APIs and technology standards for compliance (leaves).

“Gaia-X may seem gigantic but we don’t think that the big issues we are trying to tackle can be ‘sliced’ into smaller parts. In Financial Services and Insurance, this is especially true at a time when cloud adoption needs to accelerate and there is so much change within the industry.”

We will be looking out in the future to see how GAIA-X and its pilots will develop thereby changing the European Financial Service & Insurance industry. Stay tuned for more in the future.

Jonas Geisen, ESR


[1] Boley, H., & Chang, E. (2007, February). Digital ecosystems: Principles and semantics. In 2007 Inaugural IEEE-IES Digital EcoSystems and Technologies Conference (pp. 398-403). IEEE.

[2] Adner, R. (2017). Ecosystem as structure: An actionable construct for strategy. Journal of management, 43(1), 39-58.

[3] Tan, F. T., Ondrus, J., Tan, B., & Oh, J. (2020). Digital transformation of business ecosystems: Evidence from the Korean pop industry. Information Systems Journal, 30(5), 866-898.

[4] Appleyard, M. M., & Chesbrough, H. W. (2017). The dynamics of open strategy: from adoption to reversion. Long Range Planning, 50(3), 310-321.

[5] Chesbrough, H. W., & Appleyard, M. M. (2007). Open innovation and strategy. California management review, 50(1), 57-76.

[6] Hautz, J., Seidl, D., & Whittington, R. (2017). Open strategy: Dimensions, dilemmas, dynamics Long Range Planning, 50(3):298-309

[7] Chesbrough, H., Heaton, S., & Mei, L. (2020). Open innovation with Chinese characteristics: a dynamic capabilities perspective. R&D Management.

[8] Gooyert, V. D., Rouwette, E. A. J. A., & van Kranenburg, H. L. (2019). Interorganizational strategizing.

Enabling next-generation customer insights interactions in insurance through explainable AI – A session of the FINDER inclusive digital innovation week

In the last session of the FINDER inclusive digital innovation week, Jeremie Abiteboul, Chief Technology Advisor at DreamQuark, explained how explainable AI works, its benefits, and how DreamQuark is implementing it with customers.

What is explainable AI?

Explainable AI refers to making the decision-making process of a machine-learning model transparent and understandable for a human observer. This includes which data has been used as an input and which variables are proportionally contributing to a model’s decision.

Why do we need explainable AI?

There are multiple reasons why explainable is needed. Firstly, we need to know if input data is biased because that leads to bias-reproducing AI. Secondly, we need to know which variables the model is attributing the most weight to since these could be variables that discriminate against particular groups of people. Thirdly, having an explainable AI model enables companies to address accountability and to be prepared for regulatory reporting.

How to implement explainable AI in insurance?

The prominent business cases that AI in insurance addresses are cross-selling and up-selling, targeted recommendations, and churn prevention. Explainable AI in insurance (compared to non-explainable AI) enables customers to have increased trust in the AI system, validate the business relevance of the model, discover new insights in the data, check for variables that should be excluded, and use it for regulatory purposes.

Contact

If you would like to learn more about explainable AI in insurance, please reach out to Jeremie Abiteboul.

Inclusive Digital Innovation Event – Session 3 Summary

On Wednesday, March 17th, S. James Ellis gave a talk concerning ecosystem dominance at the weeklong Inclusive Digital Innovation event hosted by Atos. This discussion comes in the wake of a white paper currently in development centered around the same topic.

The paper views dominance through three different lenses in order to prescribe what incumbent and startups should focus on to gain a dominant edge in digital, data-driven ecosystems. “Dominance,” in this sense, is given a fair amount of room for interpretation, but it hinges on the idea that in an ecosystem where a business’ stakeholders seek sustainable revenue going forward, there exists the possibility to adapt to ecosystem changes while simultaneously gaining some measure of influence over how a company’s peers in an ecosystem engage with each other. This all centers on a core tenet of ecosystems being the variety of interactions between members.

Customer Access

The first argument asserts that customer access – distinct from customer engagements – is a path to focus on when seeking a dominant position in ecosystems. While many companies do indeed prioritize interaction with their customers as a general objective, this point of view suggests that building the material or conceptual infrastructure to own engagement with the customer is key to gaining a dominant advantage. This could be actualized, for instance, through building “vessel offerings,” where the focal company bundles its own offerings alongside complementary companies’ offerings. The example James gave was that of Internet companies that bundle television companies’ offerings in with their own services, thereby owning access to the Internet and television customer. As the customer, in this perspective, is assumed to be the leading force in ecosystem innovation, this begets an advantage in seizing customer-led innovation opportunities – and thus, a sense of dominance concerning this.

Resources

Similar to hallmark resource-based approaches, this viewpoint asserts that access to key resources is the key to finding a dominant position in ecosystems. However and somewhat particular to data-driven ecosystems, these key resources are interrelated proportionately. That is, a company must achieve an interlinked balance of capital, talent, and data in order to most effectively advance its position in its ecosystem. This viewpoint further posits that an overage of any of these resources without a correlated gain in the other two will result in an inefficient operating position, which could slow the company down enough to jeopardize its dominant advantage.

Tri-axis model of key resources, with the cone representing an optimal balance through growth and time.

Ecosystem Centrality

The final viewpoint asserts that a company that systematically pursues the most ecosystem connections, thus centralizing itself among participants, stands to gain a dominant edge among peers. By establishing material linkages with other companies, such as supply chain redundancies, formalized partnerships in joint offerings, and the like, this central and centralizing company begins to insulate itself from the inevitable failures and disruptions that occur in ecosystems, and especially those experiencing the turbulence of broadscale innovation.

The white paper will be available through Atos’ Thought Leadership publications later this year.

Other sessions in the event were given on de-risking corporate startups by Josemaria Siota, GAIA-X by Hubert Tardieu, life-fulfillment services as can be offered by retail banks by Eddy Claessens, and enabling next generation customer insights and interactions through explainable AI by Jeremie Abiteboul.