When there's a question, Gi has the answer. Gi is the conversational AI assistant employees and managers talk to every day, grounded in your own company knowledge. Behind Gi works a team of four specialised AI Agents that take on the heavy lifting of HR and training. Your organisation's data stays in your organisation.

Gi answers questions about company knowledge, policies, contacts, training and platform use. It learns from the documents, policies and content your organisation defines, and answers based on what your company has set. It does not invent answers.

When a document is the basis for an answer, Gi cites the source so users can read the full context. When Gi does not have enough information, it says so instead of guessing.

Gi is one assistant with a team of digital specialists working 24/7. Each agent has a clear role:

Each agent has its own help article in this collection with the detail of what it does and how to use it.

Employees: find information about the company mission, values, career progression, compliance policies, benefits, and other common internal questions.

Managers: locate information for talent management, internal mobility, and other people-related decisions.

Gi draws on two sources of knowledge inside your GFoundry environment:

Each document or content item is assigned to a single knowledge area based on its dominant theme:

Gi identifies the topic of each question and looks for the answer in the most relevant area. The result is more precise answers and clear control for admins over what each area knows.

Because each document lives in a single area, a long master document that mixes several topics will be classified by its dominant theme. Splitting the content into focused, single-topic documents helps Gi reach the right answer reliably.

Gi can adjust its response based on the employee profile available to the system. The tone and suggestions vary according to the user's stage in the company. For example, a new employee asking about career growth gets onboarding, training and integration steps; a long-tenured employee gets practical actions, well-being resources, or internal mobility opportunities.

Personalisation uses non-identifying attributes (such as tenure band or role family). The AI provider never receives the employee's name or email.

The way you organise your knowledge has a direct impact on the quality and consistency of Gi's answers.

Prefer short, focused documents over long master files that mix several themes. A document called Vacation and Attendance FAQ will outperform the same content embedded inside a generic Employee FAQ. As a working split for a typical employee FAQ, consider:

A title that names the topic explicitly (for example, Vacation Policy 2026) is easier for Gi to route than a generic one (for example, Document 14).

Lead each document with a short summary of its main topic, then go into the detail. This is especially important for PDFs.

Keep one document per language rather than mixing translations in the same file.

Restrict documents to groups only when the content is genuinely confidential or role-specific. Over-restricting reduces the pool of knowledge available to most users.

Outdated documents stay in the knowledge base and can compete with the current source of truth. Delete or update them as part of your normal knowledge-management routine.

Conversations with Gi are stored inside the platform to improve answer quality and to identify gaps in the knowledge base. Administrators can review individual chat sessions in the backoffice as a content-quality tool, not as a surveillance tool. Employees should be made aware of this through a short notice at the chat entry point.

The AI provider (OpenAI API, under a contractual no-training clause) receives the user's question, the relevant document excerpts, and a pseudonymous identifier. It does not receive the user's name, email, or manager name. The provider is contractually prevented from using prompts or responses to train its models.

Group-based document visibility is enforced before any answer is generated. A document that is restricted to a specific group is never used to answer a question from a user outside that group.

All Gi data, including conversations and the document corpus, stays inside the European Union.

Chat history retention is configured per tenant. Employees can request deletion of their chat history through their administrator. The full Data Processing Agreement is available on request.

Admins have a dedicated area in the backoffice to keep Gi's source content up to date and to monitor how people are interacting with it.

Create a new document with New Document, then give it a name, upload a file, choose a visibility, and assign a category. Documents can be set to public or private. Public documents are visible to all users; private documents are still used by Gi to generate answers, but users cannot see them in the help center. You can also segment a document by groups.

The chat sessions area lists conversations with the first user message, the date, and the number of interactions. Open View Details to read the full chat history.

Each document has a history section that records changes such as file updates, category changes, status updates, and group segmentation changes.

Rephrasing changes how the question ranks against the available documents. If you see this on a recurring topic, the source document is usually mixed with other themes and would benefit from being split into a focused single-topic document.

No. Each document is classified into a single area based on its dominant theme. If a document covers multiple topics, split it into separate single-topic documents.

Uploaded documents are added by admins through the backoffice and typically contain reference material such as manuals, policies, and FAQs. Platform content is the editorial content already published on your GFoundry platform; Gi picks it up automatically. Both are subject to the same group visibility rules.

Documents uploaded through the backoffice are available shortly after the indexing job finishes. Editorial content published on the platform is picked up automatically within a day.

The next time Gi is asked a question that retrieves that document, it will use the updated version. There is no need to clear conversations or restart Gi.

Yes. When the answer is grounded in a document, Gi includes a link to the source. Restricted documents are not surfaced to users who do not have access.

Within a chat session, Gi uses earlier turns as context. Across sessions, conversation history is stored for review by admins and to improve answer quality, but it is not used to retrain the underlying AI model.

Treat it as a signal that the knowledge base needs work. The most common causes are an outdated document still present in the corpus, two documents disagreeing on the same topic, or a master document where the relevant section is buried under unrelated content.

Open a support ticket including the chat session URL (or the timestamp of the conversation), the question, the answer received, and the answer expected. The support team can then trace which documents were retrieved and recommend the right correction.