Imago Platform Ltd. has integrated graph database technology to enhance the accuracy of its AI-driven knowledge management systems. The company initially struggled to achieve consistent results when processing complex warranty documents for its first client, a lighting and electrical business, due to the probabilistic nature of language models.
By joining the Neo4j Startup Program, Imago implemented graph databases to provide context to its AI agents. This approach allows the system to retrieve information based on established relationships between data sources, reducing hallucinations and ensuring deterministic responses. The platform now connects CRM and ERP systems with documents and emails.
Founded 18 months ago, the startup currently serves clients in sectors including property management, automotive retail, and electrical supply in Sweden. The platform utilizes a needs-based product engine that assists users in selecting items from hundreds of alternatives by analyzing specific requirements.
Beyond data retrieval, Imago uses Neo4j to monitor and trace the decision-making processes of its AI agents. As business logic increasingly shifts to the AI layer, the graph database creates nodes for every step taken by an agent, providing observability into operations running within Kubernetes networks.
