AI Overview
About 2 min
AI Overview
Page Guide
ScenarioUnderstand the relationship between models, chat, knowledge bases, protocol adaptation, and intelligent queries before enabling AI features.
PrerequisiteThe backend version includes `fastbee-ai`, and the Vue3 console can open AI menus.
AcceptanceModels are available, knowledge can be retrieved, chat returns answers, and protocol or data-query tasks enter the AI flow.
1. Capability Structure
2. Menu And Source Mapping
| Feature | Frontend Directory | Backend Prefix | Description |
|---|---|---|---|
| AI Chat | vue3/src/views/ai/chat | /ai/chat | User question and model response. |
| Chat Records | vue3/src/views/ai/chatRecord | /ai/chatRecord | Query and manage chat history. |
| Model Provider | vue3/src/views/ai/provider | /ai/provider | Provider, key, and availability. |
| Model Configuration | vue3/src/views/ai/model | /ai/model | Model name, type, parameter, and status. |
| Knowledge Base | vue3/src/views/ai/knowledge | /ai/knowledge | Upload, parse, version, and publish documents. |
| Protocol Adaptation | vue3/src/views/ai/protocolAdaptation | /ai/protocol/adaptation | Analyze protocol materials and generate access suggestions. |
| NL2SQL | Integrated by page | /ai/nl2sql | Convert natural language into structured query plans. |
3. Recommended Configuration Order
- Configure a model provider.
- Configure and enable a model.
- Test AI chat with a simple question.
- Create a knowledge base and upload documents.
- Parse documents, build a version, and publish it.
- Use the knowledge base in protocol adaptation or business Q&A.
- Configure NL2SQL only after data permission and semantic fields are clear.
4. Typical Scenarios
| Scenario | Required Setup | Output |
|---|---|---|
| Document Q&A | Model, knowledge base, documents, published version. | Users can ask questions about platform or project documents. |
| Protocol adaptation | Model, protocol knowledge, protocol materials. | Field explanation, access suggestions, and development notes. |
| Operations assistant | Model, platform knowledge, issue collection. | Assist troubleshooting for access, deployment, video, and operations. |
| Data query | Model, semantic model, data permission. | Query plan or SQL-like result. |
| Source-code knowledge | Model and source-code materials. | Help developers understand modules and APIs. |
5. Security Notes
- API keys and model secrets should be maintained only by administrators.
- Knowledge-base documents may contain customer data and should be desensitized when needed.
- AI-generated access suggestions must be verified with real device messages.
- NL2SQL must be constrained by data permission to avoid unauthorized queries.