Dnotitia Open-Sources AKB Platform to Accelerate Enterprise AI Knowledge Infrastructure

Dnotitia Inc. (Dnotitia), a company specialising in long-term memory AI and semiconductor-based AI infrastructure technologies, announced that it has open-sourced AKB (Agent Knowledge Base) on GitHub. AKB is a knowledge platform designed to help organisations turn scattered enterprise knowledge and agent-generated work context into AI-ready knowledge bases.

AKB integrates scattered internal documents, files, databases, and workflow records into a unified knowledge base that AI agents can access and use. By reducing the need for employees to manually search for materials, pass information between teams, or repeatedly reformat documents for AI use, AKB enables multiple teams and AI agents to work from a shared business context.

As generative AI moves beyond individual writing and search assistance into real business workflows across engineering, sales, HR, and marketing, enterprises are facing a new challenge. Critical organisational knowledge is often spread across document repositories, collaboration tools, presentations, and databases, making it difficult for AI agents to continuously understand and use the context needed for actual work.

This fragmentation creates inefficiencies. Employees often need to resend existing materials, ask colleagues for the same information, or convert documents into formats that AI systems can read. AKB is designed to reduce these workflow inefficiencies by providing an environment where AI agents can directly access and use organisational knowledge.

AKB consolidates different types of information, including documents, files, and database tables, into a single knowledge base. It applies an ontology-based structure that defines semantic relationships between documents and data, allowing AI agents to use not only individual materials but also the relationships among them.

Unlike traditional RAG-style knowledge bases that primarily make existing documents searchable, AKB is designed to continuously accumulate and organise the context generated by AI agents during work. This includes conversations, work records, decision context, and outputs created throughout business workflows.

Recent efforts such as LLM Wiki and GBrain reflect growing interest in long-term memory and knowledge stores for AI agents. AKB extends this direction into enterprise environments, with access controls by organisation, team, and role built into its core design.

For organisation-wide operation, AKB incorporates team-, role-, and project-level access control as well as user-level permission boundaries. This allows AI agents to use the business context they need while keeping sensitive information within defined permission scopes. The platform is designed to support both knowledge sharing and enterprise security requirements.

Technically, AKB supports MCP (Model Context Protocol)-based integration and standard Markdown-based document management. It can also manage various types of content, including SQL databases and object storage. Through graph-based relationship definitions, AKB connects relationships across work knowledge. Combined with search powered by Seahorse, Dnotitia's vector database, AKB enables users to explore business context and related information that may be difficult to find through keyword search alone.

Dnotitia has made AKB available as an open-source project on GitHub, with free use for non-commercial purposes. The company plans to continue improving the platform based on feedback from developers and business users, with a focus on capabilities needed for AI agent-based work environments.

"Enterprise AI competitiveness is increasingly shifting from which models a company adopts to how effectively AI can use the data and knowledge the organisation already has," said MK Chung, CEO of Dnotitia. "By open-sourcing AKB, Dnotitia aims to help more organisations turn their internal knowledge into AI-ready assets and grow alongside AI agents."