COMPETITIVE BRIEFING · SEPTEMBER 2026
The market is converging on organisational context.
Orient is building the layer beyond it.
Microsoft, Atlassian, Glean, ServiceNow and OpenAI are building systems that give AI better access to enterprise context, memory and relationships. That validates the underlying problem.
Orient maintains an explicit, persistent model of what the organisation knows, believes, has decided, is trying to achieve and remains uncertain about—and revises that model when new evidence warrants a change.
Others increasingly give AI context.
Orient maintains organisational state.
01 / THE ARCHITECTURAL DISTINCTION
Context is assembled. State persists.
The distinction is what happens between one run and the next. Context and memory are becoming infrastructure. Orient’s focus is the governed maintenance of organisational understanding itself.
A context-led workflow
01
Information arrives
02
Search, retrieve and connect
03
Assemble relevant context
04
A model reasons
05
Answer or act
06
Assemble context for the next task
Orient
01
Information arrives
02
Observe and interpret
03
Compare with current reality and intent
04
Judge whether understanding should change
05
Propose and ratify when required
06
Persist the accepted state
↻ Future reasoning starts from this maintained state.
Stop reconstructing the organisation on every run.
This compares architectural emphasis, not an exhaustive account of each product. Enterprise platforms increasingly persist context and memory; Orient’s distinction must be demonstrated in explicit state, governed transitions and the history connecting them.
02 / WHAT ORIENT DOES DIFFERENTLY
An observation becomes part of a continuing understanding.
A document is added. A source is watched. An article arrives. A person or agent creates something. Orient treats it as an observation to understand in relation to what the organisation already believes and is trying to achieve.
01 · Observe
Identify claims, evidence, questions, decisions and provenance. Keep the source attached.
02 · Reconcile
Retrieve relevant beliefs and intent. Determine what the observation supports, contradicts, qualifies or leaves unchanged.
03 · Govern Change
Propose a revision when warranted. Consequential changes require the appropriate authority and human ratification.
04 · Persist
Record accepted state and why it changed. Later observations reason against it, keeping a history of organisational cognition.
03 / THE PRODUCT TEST
Four questions that make the distinction concrete.
Before
What did the organisation believe before this arrived?
Change
Did this evidence change that belief?
Reason
Why—and under whose authority?
Now
What does the organisation believe now?
Search can retrieve the evidence. A context graph can connect it. A model can reason about it. Orient maintains the answer through time.
04 / THE COMPETITIVE MAP
A converging market.
A distinct state-management problem.
WHAT THEY ARE BUILDING
Semantic workplace context for Copilot and agents.
WHERE ORIENT GOES FURTHER
Maintained organisational understanding across time.
WHAT THEY ARE BUILDING
A living graph of work, people, goals and decisions.
WHERE ORIENT GOES FURTHER
Explicit beliefs, evidence, uncertainty and state transitions.
WHAT THEY ARE BUILDING
Enterprise context and an organisational graph for AI.
WHERE ORIENT GOES FURTHER
Reconciliation of new evidence against accepted state.
WHAT THEY ARE BUILDING
An operational ontology of objects, relationships and actions.
WHERE ORIENT GOES FURTHER
A model of what the organisation understands and how that understanding changes.
WHAT THEY ARE BUILDING
Governed context and operational execution for agents.
WHERE ORIENT GOES FURTHER
Governed organisational understanding before execution.
WHAT THEY ARE BUILDING
Business context, institutional memory and agent execution.
WHERE ORIENT GOES FURTHER
Model-independent, canonical organisational state.
They are building better context for intelligence.
Orient is building the state that intelligence reasons from.
05 / WHY THIS BECOMES DEFENSIBLE
A state model. A transition engine. A compounding asset.
The investment case rests on making these capabilities work together—and demonstrating that the maintained state becomes more useful as the organisation changes.
A New State Model
Claims, evidence, beliefs, questions, contradictions, decisions, expectations, situations and authority become explicit computational objects. The system represents what the organisation understands, not just where information is stored.
A State-Transition Engine
New evidence is resolved against existing propositions, evaluated and reconciled. The result may preserve the current state, open a question or propose a revision. Accepted changes remain inspectable.
A Compounding Organisational Asset
What we believed → what happened → what changed → why it changed → who ratified it → what followed. That accepted history can grow in value and become increasingly difficult to reproduce elsewhere.
Model Independence
Organisational state survives a change of model. GPT, Claude, Gemini, Mistral and specialist agents can reason from the same maintained state. Frontier models can supply intelligence to Orient; replacing a model need not mean reconstructing the organisation.
06 / THE CLOSEST ARCHITECTURAL COMPARISON
The operational world and our understanding of it.
Palantir · Operational World Model
Factory
Supplier
Aircraft
Contract
Inventory
Action
Orient · Organisational Understanding
Belief
Evidence
Question
Decision
Expectation
Contradiction
Palantir models what exists and what an organisation can do. Orient models what the organisation understands to be true and how that understanding changes.
The operational world changes. Orient maintains the organisation’s understanding of that world.
07 / WHERE ORIENT DOES NOT COMPETE
Intelligence, applications and agents can surround the same shared state.
Orient is not trying to replace enterprise search, foundation models, collaboration systems or agent platforms. It can consume them and serve them. The bet is that organisations of humans and increasingly autonomous agents need an independent, maintained state beneath those systems.
Collaboration platforms can provide observations. Models can interpret and reason. Agents can consume maintained state and return new evidence. MCP provides a way to connect those participants to the same organisational understanding.
08 / OPTIONAL DEEPER READING
Detailed competitive analysis.
The distinctions below describe Orient’s architectural position. The product must make them visible through maintained state, evidence, authority and a history of change.
Microsoft Work IQ
Work IQ moves beyond document access toward a continuously updated semantic understanding of how an organisation operates, drawing on email, meetings, chats, files and collaboration signals. If Orient were enterprise search or an AI assistant, Microsoft would already address much of the proposition. Orient is solving a different state-management problem: maintaining an inspectable record of what the organisation understands, why it believes it and how accepted understanding changes between interactions.
Atlassian Teamwork Graph
Teamwork Graph connects work, people, tools, goals and decisions in a living map. The distinction is not that Atlassian has documents while Orient understands them. It is the lifecycle of understanding: a belief may be proposed, supported, contradicted, ratified or superseded, then retrieved when later evidence arrives. Orient’s opportunity is to govern that lifecycle as persistent organisational state.
Glean
Glean’s Enterprise Graph represents relationships between people, teams, projects, processes and information, supporting both search and agents. It gives intelligence richer enterprise context. Orient focuses on what the organisation currently understands: given new evidence, what should remain accepted, what should change, what remains unresolved and why?
Palantir
Foundry’s Ontology connects enterprise data to real-world objects, relationships, actions and operational decisions. This makes Palantir a serious architectural comparison, particularly in operational environments. Orient’s emphasis is the organisation’s understanding of that world: beliefs, evidence, uncertainty, questions, decisions and the governed transitions between them.
ServiceNow
Workflow Data Fabric, Knowledge Graph, Context Engine and agent governance bring connected data, context and controlled action together. ServiceNow’s emphasis is governed operational execution. Orient’s focus is the governed organisational understanding against which execution should be assessed, including what has changed and what now deserves reconsideration.
OpenAI
Company Knowledge, Workspace Agents and Frontier bring organisational sources, business context, durable institutional memory and agent execution closer together. Orient should not compete as another AI interface. Its architectural position is a persistent organisational state that survives individual models and agents: OpenAI can supply intelligence inside Orient, while agents consume and contribute to Orient’s state.
Google Gemini Enterprise
Gemini Enterprise combines search, conversational assistance, connectors and an agent platform grounded in organisational information. Orient’s focus is the intervening state-transition system: observation, interpretation, reconciliation and governed state change before subsequent reasoning or action.
Salesforce
Agentforce and Data 360 increasingly connect context, shared understanding and agency around customer and business processes. This reinforces the shared-context direction of enterprise AI. Orient’s ambition is organisationally horizontal: a maintained understanding across functions, rather than a state layer anchored primarily in customer relationships.
Anthropic
Claude, Cowork, Claude Code, plugins and MCP make intelligence and agent execution more useful when connected to organisational tools and knowledge. Anthropic can be an intelligence and agency provider within Orient’s architecture. The distinction depends on Orient maintaining independent organisational state, rather than simply connecting a model to more sources.
Writer
Writer’s Knowledge Graph supports reliable retrieval over trusted company information. Orient’s focus extends to accepted beliefs, unresolved questions, evidence relationships and governed changes in organisational understanding through time.
Mistral
Connectors, MCP support, tool calling, approvals and an agent SDK make Mistral a possible intelligence and agent provider in the surrounding ecosystem. Orient’s proposed role is the persistent state those models and agents can reason against and help maintain.
09 / THE INVESTMENT CONCLUSION
The category is forming around us.
The largest enterprise and AI companies are converging on the same conclusion: models need organisational context, agents need memory, and enterprise AI needs a shared understanding of the business.
We believe the next step is a maintained organisational state. As humans and machines continuously observe, reason and act, the organisation needs to know what it currently understands, why, what changed and what can safely happen next.
That is the layer Orient is building.
The opportunity is to become infrastructure for organisations in which humans and machines think and act together.