Context unification technology for AI agents.
Unified Context™ is a context unification technology and standard developed by Secretar.AI that connects information, interactions and events related to the same entity across integrated systems.
Every interaction. Every system. One shared context.
A customer’s story spans conversations, documents, appointments, and business events. Unified Context™ brings related information into the context available to AI agents across integrated systems.
The entity is the reference point: a customer, for example, whose relationship with a company continues across different tools and channels.
A message arrives, a proposal is accepted, a payment is recorded, or an appointment is scheduled. Each system holds a part of the story.
Information, interactions, and events are connected to the entity they belong to, according to the sources incorporated into the integration.
An agent can interpret a new request using relevant information beyond the current conversation, alongside its tools and operating rules.
A customer discusses a proposal on WhatsApp, makes a payment, and then asks about the next step on Instagram. When the relevant systems are integrated, the proposal and payment can inform the new interaction.
The available context depends on connected sources and existing integrations. The actions an agent can take also depend on its tools, permissions, and rules.
One customer
Unified Context™ was developed by the Secretar.AI team. Secretar.AI is the first planned implementation, announced for the new version scheduled for September 8, 2026.
Read the conceptual specification on GitHub →
Draft v0.1, published for discussion. Proposed requirements and synthetic examples; not a production API or interoperability certification.
Explore the definition, product vision, launch, and practical examples in the Unified Context™ article series.
The new version of Secretar.AI arrives on September 8 with Unified Context™, connecting information from the customer journey to provide continuity in AI-powered service.
Read articleLearn the concept, the information that can make up the context, and how Unified Context™ applies to the work of AI agents across systems.
Read articleA customer’s relationship with a company spans channels, documents, and events. We created Unified Context™ to connect that context to the work of AI.
Read articleWhen information exists within the company but does not reach the person providing service, customers have to reconstruct their own story. Learn how to recognize this problem.
Read articleOmnichannel aims to provide continuity across channels. Unified context connects operational information that helps sustain this experience in AI-powered customer service.
Read articleEach stage adds information to the customer relationship. Learn which records matter and how they can support continuity in AI-powered customer service.
Read articleExplore the relationship with agent memory, RAG, context engineering, Customer 360, CDPs, omnichannel, MCP, and multiple AI agents.
Agent memory retains information across interactions. Unified Context focuses on connecting the information belonging to an entity across integrated systems.
Read articleRAG retrieves information to support generation. Unified Context addresses the connection between an entity and its information across integrated systems.
Read articleContext engineering shapes what an AI model receives. Unified Context focuses on connecting entity information that can inform that model's work.
Read articleCustomer 360 and Unified Context share an interest in connected customer information. Their useful comparison concerns scope and operational use.
Read articleOmnichannel coordinates continuity across channels. Unified Context focuses on connecting entity information across integrated systems. Their capabilities can overlap.
Read articleCDPs and Unified Context can overlap in customer data unification. The useful comparison is which component owns identity, context and access for each workflow.
Read articleMCP defines a connection protocol for AI applications and external capabilities. Unified Context addresses how information about the same entity is related.
Read articleSharing context across agents requires more than forwarding a conversation. Follow a conceptual workflow that preserves entity references, source evidence and task scope.
Read article