Published at: 2026-09-17
AI PaaS overview
ShareCRM AI PaaS is an enterprise AI capability platform. It provides end-to-end capabilities that span agent orchestration, conversational interaction, prompt assets, knowledge retrieval, and machine-learning prediction. These capabilities help frontline sales, managers, and operations staff complete complex CRM tasks through natural language.
Capability Map
AI PaaS consists of the following core capability families. You can understand and configure each capability family independently.
| Capability family | Role | Problem it solves |
|---|---|---|
| Agent Studio | Agent orchestration platform | Lets you configure, manage, and publish enterprise agents (Agents) in a visual, low-code way |
| ShareAgent | Conversational entry point and personal assistant | Provides a unified conversational workspace that discovers and invokes enterprise SubAgents to handle business requests |
| Prompt Templates | Prompt asset library | Provides reusable prompt templates with variable placeholders. It combines CRM data to generate personalized content |
| Semantic Retrieval Index | RAG knowledge index | Structures enterprise private-domain knowledge into a searchable index. It provides precise recall for RAG |
| Billing & Usage Management | Credit-based billing and usage monitoring | Uses credits as the unified unit to measure AI resource consumption. It provides cost estimation and usage viewing |
Agent Architecture Overview
ShareCRM Agent uses a “plan-execute-evaluate” Agent Loop architecture. Agents run tasks autonomously in a secure cloud sandbox. Agent Studio serves as the Setup console for orchestration. ShareAgent serves as the front-end entry point for discovery and invocation. Together they form a closed loop of “back-end orchestration plus front-end conversation.”
A token is the unified billing unit for AI PaaS. 1 credit = 0.10 RMB. For details, see 07-Billing & Usage Management.