Building an intranet with a knowledge graph is not an exclusively technical project. Although semantic modeling, data source integration, and artificial intelligence are central pieces, success depends on the right people participating from the start. The value proposition is clear: connect corporate knowledge that today lives scattered across emails, repositories, and management tools. But that knowledge can only become a useful graph if business, technology, and data governance teams work together.
The first profile to include is the executive sponsor. This is the person with budget and authority to prioritize the project. A graph-based intranet affects several areas: human resources, IT, operations, legal, and internal communications. Without a sponsor who connects the project to strategy, the initiative becomes a pilot with no continuity. This role must define business goals, release resources, and remove internal barriers.
The product owner translates the sponsor's vision into a work plan. They define priority use cases: expert search, onboarding, document management, detection of internal capabilities, and answers to business questions. They also establish acceptance criteria and decide what goes into each iteration. It is advisable to choose someone with cross-functional vision and the ability to make quick decisions.
Business users provide the real context. They know what questions they ask every day, what information they cannot find, and which relationships matter. Their participation is essential in design workshops and usability testing. Instead of assuming that the graph must contain all company data, users help prioritize the entities and connections that generate the most value. They also take part in validating the ontology so that terms have a common meaning.
The IT department participates in technical integration. The intranet needs to connect with Active Directory, collaboration tools, CRMs, ERPs, and document repositories. The technical team defines authentication models, perimeter security, backups, and deployment strategies on cloud AWS or Azure. In environments where sensitive data is combined with external interfaces, cybersecurity is critical. IT participation is not limited to the start; it includes daily operations and performance monitoring.
The compliance or risk area must get involved when the intranet handles personal data, confidential information, or regulated processes. A knowledge graph, by relating concepts, can reveal information that was not evident in separate systems. Access and visibility policies need to be defined. Early involvement of compliance avoids costly redesigns. In addition, AI operations should include human oversight mechanisms.
A role that is often forgotten is the data owner or steward. This person knows the origin of each source, quality levels, and associated business rules. The graph is built from information coming from multiple systems: billing, HR, projects, support. Without someone to validate accuracy, the graph will contain duplicates, inconsistencies, and erroneous relationships. Data stewards also define taxonomy, synonyms, and hierarchies.
You should also include a security profile from the design stage. Managing basic permissions is not enough. A graph-based intranet often serves real-time queries, combines internal and external data, and is accessible from different devices. It is necessary to review architecture, encryption, access control, and vulnerability management. Cybersecurity must be present in infrastructure and development decisions. Q2BSTUDIO includes security engineers in the architecture and testing phases.
Internal communication is key to getting people to use the intranet. The change management lead designs the training plan and measures adoption. An intranet with a knowledge graph represents a new way of working. If employees do not understand how to ask questions or interpret results, the system will be underused. It is important to include internal ambassadors who share real use cases.
From an implementation standpoint, Q2BSTUDIO recommends building a multidisciplinary team and relying on partners experienced in artificial intelligence and custom software development. The company participates in all phases: discovery, graph design, integration, deployment, and evolution. Its teams combine software architects, data engineers, AWS/Azure cloud specialists, and cybersecurity experts. This approach reduces risk and accelerates value delivery.
In daily operations, combining an intranet with a knowledge graph and BI/Power BI tools makes it possible to visualize relationships between units, skills, and workloads. In addition, with AI agents configured on top of the knowledge base, employees can receive direct answers without relying on traditional search. These agents should be trained with the company's vocabulary and supervised to maintain accuracy.
You should also define who participates in graph design decisions. Entities, relationships, and attributes represent the concepts that matter to the business. A common mistake is to delegate the entire definition to the technical team. Engineers know how to implement, but they need experts from each area to indicate what each term means, when a relationship is valid, and what information should be shown in each context. Collaboration between business and technology is the foundation of a sustainable semantic model.
Another relevant aspect is change management. The launch of an intranet with a knowledge graph does not end on the first day. It is necessary to measure adoption, detect areas with low usage, and adjust the interface or content. This requires a small steering committee with representatives from the main areas. That committee reviews indicators such as completed searches, useful answers, onboarding time, or reduction of repeated queries.
Training is also a fundamental component. Employees must learn to ask questions precisely, interpret discovered relationships, and correct the model when they detect errors. Training sessions should include real examples, visual guides, and support channels. The more familiar the organization is with the semantics of the graph, the more likely it will keep it updated.
In short, there is no closed list of roles because each company has its own particularities. However, aligning people is a necessary condition. An intranet with a knowledge graph is a business project supported by technology, not an IT project that defines the business. An executive sponsor, a product owner, real users, IT, security, data, and communication must work together. Q2BSTUDIO facilitates this collaboration through a prior discovery process that identifies stakeholders and decision-making flows.
