Loading…
Loading…
OPEN SOURCE • ENTERPRISE AI • KNOWLEDGE INTELLIGENCE
EKIP is an open-source, self-hostable, extensible enterprise knowledge intelligence platform for building AI-powered systems using your own documents, data, retrieval pipelines, and language models.
Enterprise knowledge is spread across documents, wikis, databases, applications, and teams. Finding the right information and turning it into useful answers can be difficult. EKIP provides an open foundation for building knowledge intelligence systems that connect organizational information with modern AI. You can run EKIP with your own infrastructure, connect your preferred data sources and AI models, and adapt the platform to your organization's requirements.
EKIP is built as an open-source project so developers and organizations can inspect, extend, integrate, and self-host the platform.
Run EKIP in your own environment and connect it to the systems, data sources, models, and infrastructure that fit your requirements.
Build custom connectors, retrieval strategies, workflows, integrations, and AI experiences on top of the platform.
Your enterprise knowledge can remain within infrastructure controlled by your organization. Deployment and data-management choices remain with you.
Open source makes it possible for developers to contribute improvements, integrations, fixes, and new capabilities.
Create intelligent search experiences that understand the meaning and context behind enterprise information.
Build AI assistants that retrieve relevant organizational knowledge before generating answers.
Process and organize enterprise documents so their information can become available to downstream AI workflows.
Build retrieval-augmented generation applications using your own knowledge sources and preferred language models.
Create organization-specific interfaces for employees to discover and interact with internal knowledge.
Adapt the platform for specific business domains, workflows, and knowledge requirements.
EKIP is designed around composable knowledge intelligence components.
Connect documents and other supported information sources to the knowledge pipeline.
Extract, normalize, chunk, enrich, and prepare information for retrieval.
Use embeddings and vector search to identify relevant information based on meaning rather than simple keyword matching.
Combine retrieved enterprise context with language models to produce grounded AI responses.
Design integrations so organizations can work with the AI models and providers appropriate for their environment.
Build search interfaces, assistants, internal portals, workflows, and domain-specific AI applications on top of the platform.
EKIP is intended to be more than a ready-made interface. It provides a foundation that engineering teams can adapt to their own architecture and requirements. Developers can extend the platform through integrations, APIs, retrieval strategies, AI workflows, user interfaces, and deployment configurations.
Organizations can deploy EKIP within infrastructure appropriate to their requirements. Possible deployment models can include local development environments, private infrastructure, cloud environments, containerized deployments, and organization-controlled environments. The exact deployment architecture depends on the organization's security, infrastructure, data, and operational requirements.
EKIP follows an open-source core model. The core platform is intended to remain available to the community, while Codexis Lab offers optional premium support as a commercial service for organizations that need additional assistance. Premium support does not change the open-source nature of the core project.
Avoid building every knowledge-intelligence component from scratch.
Choose your infrastructure, data sources, models, integrations, and deployment approach.
Extend the platform around your organization's requirements.
Turn disconnected organizational information into a foundation for search, retrieval, assistants, and AI applications.
The exact technology choices may evolve as the open-source project develops.
EKIP is intended to grow as an open-source project. Developers can contribute through:
Codexis Lab provides premium engineering and support services for organizations that want help deploying, integrating, extending, or operating EKIP.
Explore EKIP, run it in your environment, extend it for your requirements, and work with Codexis Lab when you need additional engineering support.