Home / Products
TUEO Product Overview · Intelligent Foundation for Enterprise IT

Ten products, one foundation

A task-driven system: from identity, collaboration, orchestration, and coding, to execution, capability management, and on-premises delivery — each product plays its own role, and all share a single identity and audit system, together forming the intelligent foundation for enterprise IT in the AI era.

Co
Experience layer · Collaboration core

Cohort Collaboration

AI employees work alongside people in the company's group chats: taking on tasks, proactively consolidating information, dividing work by role, and hosting the weekly meeting. This is the primary workspace where people and AI employees work together.

  • Group collaboration and real-time messaging
  • Proactive summaries, risk alerts, and information consolidation
  • Weekly meeting hosting with automatic meeting notes
  • Desktop client and server, every action auditable
Learn more about Cohort →
TUEO Cohort — Enterprise Collaboration Client
L
Search
G
Major deal group
Mingcha: contract review complete…
AI
AI Secretary
Today's to-dos, organized
Major deal group4 people · 5 AI employees
Mi
Mingcha · Contract AI · 09:22
Contract review complete — 3 risk clauses identified.
!Pending approvalExpires in 1:59
Urgent procurement · Servo motors ×120
¥860K · Over budget
RejectConfirm
Describe a task, @ to collaborate…
TUEO Studio · Orchestration Canvas
Add node
Trigger
LLM
Branch
Action
Skill
Trigger · Entry
Receive order
LLM · Reasoning
AI compliance review
Branch · Routing
Manual approval (over budget)
Action · Execute
Post the entry
100%|+ Add node|Run
St
Experience layer · Application orchestration platform

Studio Workflow orchestration

Without writing code, business users configure approval, fulfillment, and other workflows into applications that AI executes automatically on business data; once published, they can be called by people, AI employees, or external systems.

  • Visual configuration, no code required
  • Free orchestration of nodes such as manual approval and conditional branches
  • Publishing and version management, adjustable at runtime
  • Observable runtime, every step traceable
Learn more about Studio →
Ca
Capability layer · Capability marketplace

Capabilities Capability management

Centrally manages every capability AI employees can call: procurement, licensing, metering, and credential custody. Enterprises can assign capabilities by department or person, with clear visibility into usage and cost.

  • Capability procurement, listing, and unified management
  • Fine-grained authorization by department or person
  • Usage and cost statistics, attributable to individuals
  • Centralized, encrypted custody of external system credentials
Learn more about Capabilities →
TUEO Capabilities · My Entitlements
My entitlements Available scope Usage Orders
3
entitlements
¥12,400
total spend
1
expiring soon
CapabilityTypeExpiryThis month
DocDocument processingPerpetual licensePerpetual620
AnaData analysisSubscriptionIn 90 days410
RepReport generationTrialIn 14 days174
TUEO Executor · Tasks
Succeeded
128
Running
3
Pending review
2
Recent tasksView all →
#E-217System integration · Order completeness API● Running
#E-216Contract element extraction✓ Succeeded
#E-214Dashboard generation✓ Succeeded
All executed in isolated sandboxes · offline · fully audited
Ex
Execution layer · Execution runtime

Executor Task execution

The actual execution engine for AI employees. Operations such as writing programs, integrating systems, and calling external services all run in isolated, controlled environments — high-risk operations run offline, with a complete audit trail.

  • Isolated, controlled environments; operations never interfere with each other
  • Development, testing, and verification completed in one flow
  • Integration with existing enterprise systems and external services
  • High-risk operations run offline, fully auditable
Learn more about Executor →
Cb
Execution layer · Coding engine

Codex-build Coding execution

TUEO's coding core: AI employees read code, modify files, run tests, and open MRs in real repositories — producing real files and real diffs, not an AI's self-reported claim that it "did the work." Enterprise identity boundaries, credentials that never leave the server, and subagents and permissions verifiable at every step.

  • Real coding execution, producing real, diff-able, revertable changes
  • Managed runtime: long sessions, real-time events, permission approval prompts
  • Enterprise identity boundaries + credentials never leave the server; model keys stay server-side
  • Folder Trust gate: repository hooks require explicit confirmation before they run
Learn more about Codex-build →
Codex-build · Work Order Coding
◉ Work order #204 · Fix order completeness API
$ git clone · read · edit files
→ src/api/orders.ts +38 −12
→ test/orders.spec.ts +9 −0
$ pnpm test ✓ all passed
$ git push · open MR · update work order
Real diffVisual self-verificationCredentials stay server-side
Id
Foundation · Identity and organization

Identity Identity and permissions

The unified identity and organization foundation across the platform: log in once to enter every product, centrally manage org structure, members, and permissions, and connect to your existing identity systems.

  • Unified single sign-on (SSO) — log in once for the entire platform
  • Unified management of org structure, members, and AI employees
  • Tiered permissions, precise down to department and role
  • Integration with existing identity systems and multi-factor authentication
Learn more about Identity →
TUEO Identity · Organization and Members
◐Unified login · Single sign-on SSOEnabled
MemberDepartmentRole
YuWang YueSalesLead
MiMingchaR&DAI employee
YzYunzhongProcurementAI employee
TUEO SDK · Plugin Development
Plugin files
manifest.ts
backend.ts
frontend.tsx
— Define once
— Used by AI as a tool
— Used by people as a UI
export default defineCapability({
  name: 'Contract risk detection',
  input: { file: File },
  output: { risks: Clause[] },
  async handler(ctx) { … }
})
Sk
Ecosystem · Developers and extensions

Skillplugins Extension development

An extension framework for developers. Enterprises can turn their own systems or specialized capabilities into plugins for AI employees to call and for business users to use — define once, use it two ways.

  • Standard plugin development framework and SDK
  • Define once — used by AI as a tool and by people as a UI
  • Secure integration with your own systems and data
  • Plugins run in isolated environments, without interfering with each other
Learn more about Skillplugins →
Cs
Foundation · On-premises delivery

Console On-premises delivery and operations

The foundation that delivers and runs the entire system in your own environment: offline installation, disconnected operation, and a unified operations console for managing configuration, logs, status, and deployment.

  • Offline packaged delivery, deployable fully disconnected
  • All-in-one installation, independent of existing infrastructure
  • Unified operations console: configuration, logs, and status centrally managed
  • All data stays inside the enterprise; nothing leaves
Learn more about Console →
TUEO Console · Operations Hub
Enterprise internal networkAll services operational
Identity
Identity · CPU 12%
Cohort
Collaboration · CPU 34%
Studio
Orchestration · CPU 18%
Executor
Execution · CPU 27%
Config · Logs · Deploy managed in one place✕ Data stays in-house
En
Foundation · Knowledge & memory substrate

Engram Knowledge & memory substrate

TUEO's suite-wide knowledge / memory foundation, a foundation that sits alongside identity: document knowledge base, conversation memory, and entity relationship graph in one — auto-organizing and self-growing. Private, lightweight, data never leaves your domain.

  • Document RAG + conversation memory + entity graph, three in one foundation
  • Dream auto-organization: the knowledge base grows itself, contradictions resolved and noise removed
  • Content-safety gating + sensitivity tiers, more configurable the more sensitive it is
  • Lightweight single-Postgres foundation, data stays in-domain, Chinese-first
Learn more about Engram →
TUEO Engram · Knowledge & Memory Substrate
Knowledge base · Memory · GraphAuto-organization running
Document knowledge base
1,284 docs · hybrid retrieval
Conversation memory
3 sessions distilled
Entity relationships
256 entities · multi-hop
Content safety
3 sensitivity tiers · gated
Data stays in-domain · domestic models✕ Knowledge compounds
In
Foundation · Compute & model serving

Inference Compute & model serving

A unified model-serving and compute control plane for your private enterprise AI infrastructure: heterogeneous accelerator cards managed independently, model serving, training, access governance, usage, and audit — all verifiable, observable, and rollback-able in one control plane.

  • Heterogeneous compute managed independently, never disguising different hardware as one pool
  • OpenAI-compatible API that hides underlying runtime differences
  • Access governance + quotas + usage + audit, one closed loop tied into identity
  • Offline install, upgrade, rollback, diagnostics, and on-site acceptance
Learn more about Inference →
TUEO Inference · Model-Serving Control Plane
Dual-pool compute · Model servingServices operational
Ascend pool
24×910B3 · GPU 61%
Hygon pool
8×DCU · GPU 43%
Model serving
OpenAI-compatible · streaming
Training jobs
LoRA SFT · queued
Quota · Usage · Audit in one loop✕ Two pools, independent

Curious how these products work together?

One real example: how the ten products close the loop on a single order in a week.

View how it works → Book a demo