RAG Mode｜Enterprise Knowledge
Find answers in your documents, with sources you can check.
Turn policies and procedures into cited answers your team can verify.
WinChat Enterprise AI workspace
Connect your knowledge
to an AI Agent
Connect your documents and business systems.
Let AI agents answer questions and take action.
Plans and actsAI agents
Approved toolsTool use
Existing workflowsConnected systems
WinChatENTERPRISE AI WORKSPACE
WHY WinChat
Build Trusted Enterprise AI Agents
Give teams trusted answers and access to the tools they need.
APPLICATION MODES
Two AI Application Modes
Choose cited Q&A or agents that can use approved tools.
AI Agent Mode｜Smart Collaboration & Decisions
Connect AI to your systems to automate everyday tasks.
Connect approved tools and live systems so agents can support daily decisions.
CAPABILITIES
Six Knowledge Capabilities
Enterprise Knowledge and Intelligent Q&A
Search documents, connect data and deliver cited answers across teams.
Doc Intelligence
Extract key information from unstructured content in many document formats.
No-Code Tools
An intuitive web interface for managing resources, permissions and projects.
Multiple KBs
Create dedicated knowledge bases for each team, with support for many data types.
Grounded AI
Back answers with key citations and traceable content to sharply reduce hallucination risk.
Integrations
An easy-to-use Web API supports flexible expansion and custom integration.
Smart Retrieval
Combine semantic search, hybrid search and GraphRAG to find exactly the knowledge you need.
USE CASES
Enterprise AI Agent Applications
Search company knowledge and let AI agents support daily work.
Customer Support KB
Find standard answers quickly so customer service replies stay consistent.
Find standard answers quickly so customer service replies stay consistent.
Look up the right procedures instantly and reduce errors.
Find the legal basis, with citations you can trace.
AI agents help new hires learn faster and get up to speed sooner.
Build dashboards in natural language and monitor key operations in real time.
FEATURES
AI Agents That Work With Your Systems
Connect knowledge, tools and business systems in one workspace.
Chat Workspace
Switch between RAG and Agent modes, ask about documents and revisit past conversations.
Review standard toolkits and system tools, switch each AI skill on or off, and set tool parameters to control access.
Build dashboards for alerts, throughput and operating trends.
Organize PDF, Word and Excel content in searchable knowledge bases.
Add MySQL, PostgreSQL or CSV connections, test them, then let AI query the live data sources you approve.
Create and manage API keys, set RAG or Agent mode and knowledge base access, and monitor gateway health and usage.
WinChat WORKFLOW
How WinChat Works
Add knowledge, set access and start asking.
- 01Set Up Knowledge and Data
Add documents, choose Vector, Hybrid or GraphRAG search, and connect approved data sources.Set Up Knowledge and Data - Set Skills and Permissions
- Start a Conversation
- Review Sources and Optimize
- Monitor and Integrate
INTEGRATION
Connect Knowledge to Action
WinChat × WinHub.AI Platform × Business Systems
Bring company data and systems into a secure AI workflow.
Business Documents (PDF / Word / Excel)SQL Databases (MySQL / PostgreSQL)CSV Files
Knowledge Base
Vector IndexHybrid IndexGraphRAGDocument Understanding
Chat Engine
RAG RetrievalAgent ReasoningSource CitationsMulti-Turn Chat
Skill Governance
Toolkit ManagementKnowledge / System ToolsNumOps Data AnalyticsAIWinOps Image Analysis
Data Insights
Real-Time QueriesAuto ChartsMonitoring DashboardsNatural Language Analysis
Standard HTTP APIAPI Key PermissionsRAG / Agent ModesSSE StreamingSession Control
Support KBSOP & Policy SearchCompliance SearchNew Hire TrainingOps Dashboard
01
Knowledge & Data
Business Documents (PDF / Word / Excel)
SQL Databases (MySQL / PostgreSQL)
CSV Files
02
WinChatAI Agent Workspace
Knowledge Base
- Vector Index
- Hybrid Index
- GraphRAG
- Document Understanding
Chat Engine
- RAG Retrieval
- Agent Reasoning
- Source Citations
- Multi-Turn Chat
Skill Governance
- Toolkit Management
- Knowledge / System Tools
- NumOps Data Analytics
- AIWinOps Image Analysis
Data Insights
- Real-Time Queries
- Auto Charts
- Monitoring Dashboards
- Natural Language Analysis
03
API IntegrationBusiness Systems
Standard HTTP API
API Key Permissions
RAG / Agent Modes
SSE Streaming
Session Control
04
Applications
Support KB
SOP & Policy Search
Compliance Search
New Hire Training
Ops Dashboard
DEPLOYMENT
Deployment Architecture and System Requirements
Deploy on-premises and keep your data and models under your control.
Software requirements
| Item | Version |
|---|---|
| Operating System | Ubuntu 24.04 LTS (64-bit) or later |
| NVIDIA GPU Driver | 535.154.05 or later |
| Docker Engine - Community | 20.10.17 or later |
Deployment architecture
- Flexible on-premises / private cloud deployment
- Data stays 100% on-site, under your control
- Multi-user collaboration and access control
- Hardware resource scheduling and quota management
Hardware requirements
| Item | Minimum | Recommended |
|---|---|---|
| CPU | 12 cores or more, 3.5GHz | 20 cores or more, 3.5GHz |
| GPU | 24GB VRAM (NVIDIA RTX 4090) | 96GB VRAM (NVIDIA RTX Pro 6000 Blackwell) |
| Memory | 128GB | 256GB |
| Storage | 1TB SSD + 4TB HDD | 2TB SSD + 8TB HDD |
| Supported Models | gpt-oss-20b | gpt-oss-120b or a model of similar size |
| Ports | USB 2.0 / 3.0 (for security dongle) | |
Recommended compatible hardware
PRICING
Purchasing Options
Choose the licensing model that fits your use case and budget.
Perpetual Plan
- Perpetual license
- Full GPU server configuration
- Private cloud / on-premises deployment
- Full control of your data
- Lifetime technical support
Best forLong-term use, sensitive data
Annual Plan
- Annual license
- Full GPU server configuration
- Private cloud / on-premises deployment
- Regular system updates
- Annual technical support
Best forFlexible budgets, phased needs