aikux.Brain

The heart of aikux.Brain is the Enterprise Knowledge Graph, which connects data sources, recognizes relationships, and thereby delivers context-based knowledge. Instead of isolated data, relationships emerge in a graph database.

1.1 Use of GraphRAG

The platform relies on GraphRAG (Retrieval-Augmented Generation using Knowledge Graphs). GraphRAG combines knowledge graphs and vector search to provide AI with deep context and Multi-Hop Reasoning equip.

• Connecting content: Content is not only stored, but also networked through relationships. The combination of GraphDB + LLM (Large Language Model) makes it possible to understand connections.
• Targeted queries: Through GraphRAG, the AI ​​specifically accesses relevant nodes and can thus provide answers with Context and justification This methodology is considered superior to purely vector-based RAG and is intended to become the standard architecture.

1.2 The Engine: Neo4j

As a graph database, aikux.Brain uses the Neo4j Graph DatabaseThe software supports the modeling of entities and relationships as well as the Multi-Hop Reasoning about complex structures. The costs for using the Neo4j Enterprise Edition are included in the Professional and Enterprise Editions of aikux.Brain.

---

aikux.Brain transforms scattered knowledge into an engine for decisions and innovation.

2.1 Explainable answers (Explainability)

A key feature for enterprise applications is transparency.

• Auditability: Every AI response is based on Paths in the graphUsers receive answers including Source path and contextThis provides companies with the ability to understand sources, dependencies, and the reasons for AI output—there is no black box.
• Contextualization: The platform is designed to enrich answers by connecting facts across data to produce more precise and meaningful results.

2.2 Enterprise architecture and scaling

aikux.Brain is designed to meet the needs of complex organizations.

• Multi-brain architecture: Allows teams or projects to work in their own “brain” (a standalone database) that clearly defined and yet networkable The platform supports the Orchestrate complex queries in seconds.
• Data integration: The solution can connect various data sources such as CRM, ERP, chats, and databases. Semantic & LLM Layer combines search with models and integrates human feedback.

---

aikux.Brain was Data sovereignty, compliance and security developed as a basic principle.

• Deployment options: The platform is either On-premise or in a certified EU cloud This ensures that the data never leaves the EU.
• Safety: The architecture is from the beginning Enterprise security It includes Tenant separation per Brain, complete Audit logs for all interactions, and compliance with regulatory standards (e.g. GDPR, ISO 27001, SOC).
• Data protection guarantee: Customer data is never used to train AI models.

---

The licensing model is Pay-per-use per brain and does not require user licenses.

• The costs consist of a basic edition fee, the DB queries (pay per use) and the passed-through AI provider costs (without surcharge).
• The Professional and Enterprise Editions include the cost of using the Neo4j Enterprise Edition.
• DB requests will be priced on a degressive basis: from €0,05 per request (1–100.000/month) to €0,01 per request (from 1.000.001/month). Thanks to the threshold effect, the new price will apply retroactively to all requests.

---