Arango Contextual Data Platform (ArangoDB)

Arango, formerly known as ArangoDB, is operated by ArangoDB GmbH in Cologne, Germany, and was co-founded by Claudius Weinberger and Dr. Frank Celler. It provides a natively multi-model database and data platform that unifies graph, vector, document, key-value and search data, aimed at enterprises that want AI agents and assistants to work from governed, connected context instead of a patchwork of separate databases.

The stack has three layers: ArangoDB as the data foundation, the Arango Platform Suite with visualization, high availability, observability, RBAC and SSO, and the Contextual Data Platform, whose 4.0 release adds AutoGraph to build knowledge graphs, AutoRAG and Deep Search to choose between GraphRAG, HybridRAG and VectorRAG, AQLizer for natural language queries and the Ada assistant. It runs self-managed on Kubernetes, VMs or bare metal, as the managed AMP cloud on AWS and GCP, or embedded via OEM licensing, with customers such as NVIDIA, HPE, London Stock Exchange, Siemens, Telstra, Orange and Stockbit.

• Multi-model engine: Graph, vector, document, key-value and search data in one database with the AQL query language.
• AutoGraph: Builds and maintains knowledge graphs automatically from enterprise data.
• AutoRAG and Deep Search: Selects GraphRAG, HybridRAG or VectorRAG and returns one governed answer.
• AQLizer and Ada: Natural language querying and an AI assistant for building and managing the platform.
• Platform Suite: High availability, observability, RBAC, SSO and audit features for production.
• Deployment choice: Self-managed on Kubernetes, managed AMP cloud on AWS and GCP, or OEM embedding.

A Vietnamese telecom operator can build a network asset knowledge graph that an AIOps assistant queries to trace outages. A bank can unify customer, transaction and document data to power an explainable internal advisory chatbot. A software company can embed the database through OEM licensing to add graph and vector features to its own product.

Teams can replace a mix of graph, vector, document and search stores with one platform, which the vendor says simplifies the stack by around 70 percent. Organizations cited in the 4.0 announcement report 2 to 4 times faster AI development cycles and 20 to 35 percent better AI decision accuracy. Graph-native lineage makes agent answers explainable and auditable, and flexible self-managed, managed and OEM options suit regulated or sovereign deployments.

Enterprises moving from AI pilots to production agents find that answer quality depends on connected, governed context, driving interest in knowledge graphs and multi-model platforms. In Southeast Asia, telecom operators, financial firms and manufacturers exploring sovereign or on-premises AI deployments need data layers that can run in their own environments.

Arango is the new company name of ArangoDB; the data controller is ArangoDB GmbH, Hohenstaufenring 43-45, Cologne, Germany. Board members include representatives of Bow Capital, IRIS and New Forge. The company is a member of the NVIDIA Inception and AWS ISV Accelerate programs and reports more than 200 production environments worldwide.