Level 4 "Really FAIR"
- Giovanni Nisato
Level | Nickname | Marketplace metaphor | Features | Picture |
4 | "Really FAIR " | "Hyper Market" | Operational, best practice known at the time of writing. Internal organizational focus. Emerging cross-company |  |
Table level 4 summary
FAIR Maturity Matrix: maturity levels (columns) | Level 4:"Really FAIR" summary
Level 4 - Capabilities
Additionally to level 3 capabilities, FAIR data is prevalent across departments and divisions, with consistent use of persistent identifiers, controlled vocabularies for key domains and business areas. Data is described in cross-domain models, and applying enterprise-level metadata standards. FAIR tools enable automated data exploration and reuse.
Leadership mandated FAIR implementation cases align with governance processes and risk management. A FAIR strategy supported by allocated resources and employee awareness is in place. Key FAIR roles are established, with internal FAIR training programs ensuring proficiency. Use cases enabled by data interoperability across organizational divisions or functions become possible.
Level 4 - Business value
Facilitated regulatory compliance and auditing. Interoperability enables data-set integration. Data exchange across functions enabled insight generation (e.g., competitive intelligence, market access, clinical development), and quality AI training data sets for LLM.
Level 4 - Questions to ask
Are the metadata “grounded” in shared vocabularies?
Does the data and metadata make relaxed use of ontologies and vocabularies that are themselves, FAIR?
How is FAIR data distributed across departments and divisions within the organization , and what are the key characteristics of this data in terms of identifiability and resolvability?
How are cross-domain analytics tools enabled within the organization?
How does leadership ensure that FAIR implementation is integral to the organization's data governance processes and risk management practices?
Describe the components of the FAIR data strategy in place, including its accommodation of both centralized enterprise-level and federated domain-specific FAIR data.
How does a Citizen Data Scientist utilize FAIR data within a pharmaceutical company?
Explain the involvement of business roles in contributing to the creation of FAIR data resources (e.g. knowledge graphs). How are these roles promoting and monitoring the value of FAIR data within the organization?
Discuss the development of organization-wide standards for FAIR data management and the integration of FAIR processes within a broader community .
How are assets and support utilized to appropriately train everyone in the organization on FAIR data?, and
How does the organization establish a registry of FAIRification tools and ensure their integration into company-level processes to support FAIR data adoption and governance?
Level 4 - FAIR data
One can find "FAIR" data across departments and divisions in an organization.
Data and Metadata are identifiable and resolvable and use Controlled Vocabulary terminologies.
Persistent Identifiers, global (GUPRI) where appropriate, are implemented consistently across the company.
In addition to the level "of "pretty FAIR", Data is described in a cross-domain data or industry model.
Data embeds Enterprise master and reference data. Enterprise-level Metadata standards are applied.
Cross-domain Analytics is enabled and showcases the "Interoperability' functionality within a given company.
Role: Citizen Data Scientist using all relevant pharma company data.
Machine interpretation is possible at an organization level.
Data starts to be "generated" FAIR.
Most of the relevant data in the company is FAIR, and there is enough knowledge and capacity (budget & competence) to make data FAIR. FAIR data is becoming the norm and part of all new projects. Making data fair is a no-brainer. It is possible and well-supported. Data that is not FAIR exists. It is intentional, e.g., based on business value, or there may be a plan to be made FAIR.
(Back to the FAIR Matrix).
Level 4 - FAIR leadership
Leadership expects that a data project (or product) budget approval will only commence with a related FAIR implementation budget.
The organization leadership has implemented a FAIR mandate across its entire enterprise, emphasizing the importance of FAIR data principles and financing internal implementation projects. This commitment is integral to their data governance processes and risk management. The focus is on internal (intra-company) leadership rather than explicitly externally oriented leadership. Nevertheless, the organization attracts vendors and service providers in the FAIR data space. It exerts external leadership, for example, by defining data and metadata requirements for service procurement in a given domain (e.g., transcriptomics or imaging).
The organization is actively involved in the broader FAIR community (outside of the organization boundaries), frequently participating in conference speakers and consistently publishing work with a foundation in FAIR principles. This activity demonstrates their dedication to promoting and advancing FAIR data practices and principles within the pharma and LifeSciences industry). The organization uses open-source tools and semantic resources to promote reuse and interoperability in a pre-competitive fashion. The accountability of leadership in FAIR implementation across the organization is clear.
(Back to the FAIR Matrix).
Level 4 - FAIR strategy
There is a FAIR data strategy in place which accommodates both centralized enterprise-level FAIR data and federated domain-specific FAIR data, and there is an operating model, set of procedures, central models, and standards to harmonize, align, and connect functional domain data to expand and scale FAIR data beyond single or simple domain-specific use cases.
The organization has set forth a comprehensive FAIR mandate across the enterprise. This mandate includes the integration of FAIR metrics into the data governance process and data risk management practices. The organization has allocated capacity (budget & people) to deliver on its FAIR data vision. The organization had the time to show the continuity of its strategy and can point to impact areas that have typically been there for several years.
There is a strategic allocation of Human, financial and time Resources to the various FAIR dimensions (cf: Tools, Roles, dimensions)
There is a clear awareness of the FAIR data strategy, especially for the roles that require its implementation. The organization's strategic culture embeds the FAIR data strategy. All employees have a basic understanding of FAIR data and act to enable its principles.
(Back to the FAIR Matrix).
Level 4 - FAIR roles
Key roles are in place for metadata, data and governance management. The additional roles that emerged more clearly include Citizen Data Scientist and Data product owner.
Leadership (C level) plays a key role and enables a data-centric enterprise, but all employees are involved implicitly (cf FAIR tools) or explicitly in the FAIR data principle implementation.
"FAIR" is not limited to technical roles. Among others, business roles provide subject matter expertise for creating Knowledge Graphs (KG) and ontologies. Financial roles ("CFO") can ensure financial support but also monitor the added value (ROI) of FAIR data, tools, and insights created by FAIR data, tools and processes.
Crucially, we see the institutionalisation of training programs ensure proficiency in each role. Change management roles and practices (cf. FAIR Processes) are well established.
The company is resourced (cf. FAIR Strategy L4) with a diverse set of roles within an existing framework for data management and integration, including, for example, Data standard experts, data curators, semantic web experts, knowledge engineers, data strategists, communication experts, enterprise and domain ontologists, AI professionals, RDF experts (NB this is not an exclusive list).
Citizen Data Scientists leverage self-service analytics platforms, intuitive data visualization tools, dashboards and other FAIR-enabled resources to explore datasets, generate reports, and uncover patterns or trends relevant to their roles.
These roles collaboratively contribute to effective data management, integrating various expertise and responsibilities.
There is enough knowledge and capacity (budget, human resources, competencies) to make data FAIR. FAIR is often the norm; all employees have heard about it and can easily access training and champions. The framework of roles and interactions is in place, along with the establishment of training for each role.
(Back to the FAIR Matrix).
Level 4 - FAIR processes
The organization has established FAIR data practices across its data ecosystem, and role-specific training is available. Employee onboarding includes FAIR data training, and this training receives regular updates. It started a journey towards "FAIR practices".
The integration of Automated FAIR data tools into the organizational system is evident. Key activities now operate with FAIR processes recognized for their value, supported by general guidelines. There's a defined set of necessary organizational processes and an awareness of existing gaps.
Feedback mechanisms exist to enhance data FAIRness, ensuring ongoing improvement—the design of Processes to maintain business metadata systematically, which is crucial for capturing usage and impact. The organization showcases the value derived from normalized FAIR data practices and continues implementation, striving towards full FAIR compliance, demonstrating a commitment to evolving and embedding FAIR principles.
Measures of business impact have been described and piloted. There may be established processes and knowledge sharing for FAIR implementation projects' business value and impact.
The definition and creation of FAIR "standards" require processes that appear at this stage. There is a dialogue beyond the boundaries (departments) of the organization.
Processes to develop organization-wide standards for FAIR data management begin to evolve.
FAIR processes are used within a broader community and are becoming standard within the organization.
(Back to the FAIR Matrix).
Level 4 - FAIR knowledge
Assets and support are in place to train everyone in the company on FAIR data. Most people in the company understand FAIR data and its value. Newly onboarded people and projects now also go through the FAIR data education.
Knowledge about validating instance metadata against annotation requirements and data shapes is evident.
Implicit knowledge with strategic value starts to be captured explicitly in (business-relevant) Knowledge Graphs.
An organization-wide cross-domain FAIR community of practice is established, with a named leader and regular cadence of meetings with a standing agenda of project reviews, best practice development, real-world experiences, and sharing of learnings.
Through organizations (such as GO FAIR FOUNDATION, ELIXIR, RDA, and the Pistoia Alliance), industry organizations have begun to come together in communities to work on FAIR standards in different domains.
The organization has gained recognition within FAIR circles. The experts within the organization regularly speak at conferences. It consistently publishes work firmly rooted in FAIR principles, highlighting its dedication to promoting the Findable, Accessible, Interoperable, and Reusable aspects of data management and contributing to thought leadership in the field.
(Back to the FAIR Matrix).
Level 4 - FAIR tools and infrastructures
The organization has taken several noteworthy steps to advance FAIR data management, including the following tools:
Automated Data Exploration Tools: The implementation of Tools enables more efficient data analysis and utilization to facilitate automated data exploration.
Defined Interaction Mechanisms: Defined Interaction mechanisms with FAIR data have been consistently implemented across the organization and meet prioritized, organization-specific business needs.
Reusability: Each dataset has a clear human and machine-readable usage licence. Data producers strive for reusability of their data to maximize their value
Organizational-Level Framework: The organization has established a comprehensive framework for FAIR data management. This framework includes a range of internal and external tools covering ontology and taxonomy sources, tools for managing ontologies and taxonomies, data governance tools, and API tools that support FAIR data adoption. Furthermore, developing strategies and tools guide the FAIRification of legacy data.
Registry of FAIRification Tools: Establishing a registry of tools specific to the organization's industry specialization. These tools are applied during the design phase before implementation and company-level processes are in place. Standards and governance practices integrate these tools, and a list of tools or internal experts with an overview of their capability is available. The organization periodically evaluates historical and current solutions, ensuring they remain current and effective.
(Back to the FAIR Matrix).
Findability:
Tool(s) or infrastructure component which contributes, enhances or enriches Findability:
Accessibility:
Tool(s) or infrastructure component(s) which contributes, enhances or enriches Accessibility.
Interoperability:
Tool(s) or infrastructure component(s) which contributes, enhances or enriches Interoperability.
Reusability:
Tool(s) or infrastructure component(s) which contributes, enhances or enriches Reusability.