NexusAI
5A product family for business understanding, AI creation, enterprise governance and organizational execution, with a dedicated AI phone service.
EDE AI TECH
Product architecture
Nine products. One architecture map.
EDE AI Tech
Start with the product families, then explore their relationships.
Product families describe the brands. The capability architecture and connections below explain how individual products divide the work.
OpenFlux
3Built around Agent execution: the desktop workspace serves personal work, Runtime uses the same engine to take organizational tasks without an interface, and Router connects team collaboration channels and endpoints.
Narriflow
1An independent product brand for presenting knowledge, interactive explanation and sharing.
The execution layer for organizational work
OpenFlux Runtime carries out the AI work you assign.
Built on the OpenFlux execution engine, OpenFlux Runtime has no user interface. It takes AI tasks assigned by NexusAI OS, works in the project environment and returns results and execution records to OS.
From Agent definitions to task delivery
Execute AI tasks and return results and records- NexusAI Studio
01Define the Agent
Set the Agent’s responsibilities, instructions and work requirements.
- NexusAI OS
02Assign the task and context
After process approval, deliver Agent settings, goals, project materials and work requirements with the task.
- OpenFlux Runtime
03Execute and return results
Complete the AI work in the project environment and return deliverables, execution records or the reason for a blocker to OS.
OS organizes goals, processes, people and decisions. Runtime registers and claims tasks, follows the Agent settings and project materials delivered with each task, works with files, commands and code in the project container and workspace, and returns deliverables, execution records or a reason it cannot continue.
Results return to the OS project
Deliverable files, code changes, service links and execution records return to the relevant project for the team to review, accept and build on. Blockers requiring a human decision also return to OS.
How Runtime works with the other products
Task assignment and results
OS defines what an AI step should do and the context it needs. Runtime claims and executes the task, then returns results, execution records or the reason for a blocker to OS.
Task requirements and project materials ↔ results and execution records
Agent definitions delivered through OS
Studio creates and configures Agents. OS references those definitions and, after process approval, delivers them with tasks to Runtime, which follows the supplied responsibilities and instructions.
Agent definitions → OS tasks → Runtime
Model access governed by Atlas
Runtime sends model requests through the enterprise model gateway. Atlas applies the relevant organization’s model access, cost budget and audit rules and records usage.
Model requests, usage and audit records
One engine, different ways to work
OpenFlux and Runtime share an execution engine. OpenFlux serves people at the desktop; Runtime has no interface and takes organizational tasks assigned by OS.
A shared engine, with separate entry points for desktop work and organizational tasks
Product relationships
Understand each product’s role, capabilities and collaboration
Follow each product from input through processing to delivery. Definition references, governance services, task delivery and message delivery are specific connections. Complementary responsibilities describe ways to divide and combine work; independent applications and capability exports each have their own use. Planned connections indicate a future direction; interfaces and collaboration remain to be defined.
Organizational decisions & execution
NexusAI OS
Its role in the ecosystem
Turn goals into ongoing work for people, Agents and projects
NexusAI OS organizes goals, cross-department processes, people and Agents in project context. It assigns and follows up work, preserves human decisions and project materials, and receives results from AI tasks carried out by Runtime.
Inputs & context
Goals, processes, people and Agent definitions
Outputs & deliverables
Tasks, decisions, progress and work artifacts
From goals and responsibilities to project results
- 01
Establish goals and requirements
Prepare organizational principles, goals, deadlines and project materials; clarify the work to advance.
- 02
Arrange processes and responsibilities
Set cross-department steps, human tasks and Agent responsibilities, and assign AI tasks to Runtime.
NexusAI OS - 03
Follow decisions and results
Review results and execution records, accept or return deliverables, and follow up waiting work and blockers.
Core capabilities
Goals and processes
- Organize work around goals
Maintain organizational principles, strategic and profit goals, metrics and deadlines. Connect goals to business processes and projects so participants understand the purpose and requirements.
- Design cross-department business processes
Arrange steps and dependencies in department swimlanes. Assign work to people, Agents or subprocesses so a project follows a defined division of responsibilities.
Human decisions and project materials
- Keep human decisions and input
Turn human work, review and questions into tasks. People can add real-world context, accept deliverables or return them for revision, with their decisions and feedback retained in the project.
- Share project materials and deliverables
Maintain key information, files and deliverables on the project so later steps can use existing materials. View documents, execution progress and time spent on the page.
Ongoing coordination and proposal comparison
- Follow up and handle blockers
The coordination center tracks progress, assigns work and follows up. It uses available information to address blockers or refers them to people, retaining execution state so work can continue after waits and interruptions.
- Compare multiple AI results
For proposal tasks, produce several results from different perspectives, then select one or combine their strengths. Keep the recommendation reasons and let people choose another result or request a new evaluation.
Product relationships
OS uses Agent responsibilities and instructions from Studio and identity and organization information from Atlas. It organizes work and assigns AI tasks to Runtime; OpenFlux provides a complementary desktop workspace.
Definition referenceNexusAI Studio
Reference NexusAI Agent definitions
OS references Agents configured in NexusAI for AI steps. Once the process is approved, those settings travel with each task to the execution endpoint.
Capabilities & deliverables involvedAgent definitions, task requirements and project context
Task deliveryOpenFlux Runtime
Hand AI tasks to OpenFlux Runtime
OS decides what an AI step should do and what it is based on; Runtime claims the task and carries it out in the project environment. OS continues once it receives results, work records or the reason the task is blocked.
Capabilities & deliverables involvedTasks, Agent settings and project materials ↔ results and work records
Governance serviceNexusAI Atlas
Use Atlas identity and organization context
OS obtains user and organization information from Atlas to assign people and responsibilities across departments. Atlas retains its identity and organization role; OS maintains goals and work state.
Capabilities & deliverables involvedUsers, departments and organizational scope
Complementary responsibilitiesOpenFlux
Organizational coordination complements desktop work
OS organizes goals, cross-department processes and human–Agent tasks. OpenFlux focuses on desktop tools, files and personal work context, addressing desktop actions alongside organizational progress.
Capabilities & deliverables involvedOrganizational coordination and desktop work capabilities
Clear boundaries
Different responsibilities. Complementary capabilities.
Business modeling, Agent building, governance, work coordination, task execution and channel delivery answer different questions.
Business semantics & Agent creation
NexusAI Terrain
Terrain models business objects, rules and actions, with evidence and human confirmation.
NexusAI Studio
Studio defines how Agents use instructions, models, knowledge and tools.
AI workflows & organizational processes
NexusAI Studio
Studio workflows arrange steps, branches and tools within an AI application.
NexusAI OS
OS coordinates work across departments and time, including people, decisions, follow-ups and artifacts.
Organizing work → executing tasks
NexusAI OS
OS decides what to do, who handles it and what it is based on, follows progress and keeps human decisions.
OpenFlux Runtime
Runtime claims AI tasks, completes them in a project environment and returns results, work records or the reason it is blocked.
Usage boundaries → work progress
NexusAI Atlas
Atlas decides who can use which applications and models, with quotas and usage records.
NexusAI OS
OS decides what happens next, who handles it and how work moves past a blocker.
Desktop workspace ↔ headless execution
OpenFlux
OpenFlux serves people at the desktop: they assign tasks, use desktop tools and review results.
OpenFlux Runtime
Runtime has no interface and takes tasks assigned by OS. Both run on the same execution engine.
Desktop execution ↔ channel delivery
OpenFlux
OpenFlux runs Agent work and manages desktop tools and project context.
OpenFlux Router
Router connects collaboration platforms, terminals and messages, delivering context and results.
How capabilities work
Follow the work through each product.
Concrete examples of what each product handles, from inputs to useful outcomes.
Turn business data into callable capabilities
Terrain preserves business meaning and evidence before exporting tools.
Business materials
Data structures · samples · SQL
Model & review
Objects · rules · processes · evidence
NexusAI TerrainCapability export
Queries · actions · MCP · Skills
Hand an AI step to the execution endpoint
OS decides what to do and what it is based on; OpenFlux Runtime does the work.
Assignment & context
Goals · Agent settings · project materials
NexusAI OSExecute in the project environment
Container · workspace · code branch
OpenFlux RuntimeResults back in the project
Deliverables · service links · work record
Bring an Agent into a team conversation
The routing service and desktop execution environment have distinct jobs.
Collaboration channels
Feishu · DingTalk · WeCom · Slack
Connect & deliver
Messages · context · terminal assignment
OpenFlux RouterExecute & respond
Agent work · tools · results
OpenFlux
Deliver a dedicated phone or knowledge experience
Choose the application around its own input and audience.
Phone service
SIM device → voice gateway → call handling
NexusAI SmartPhoneInteractive knowledge
Materials → narration → questions & sharing
Narriflow