SupportOS Platform Guide
Welcome to the SupportOS Platform Guide. This documentation explains how the system works (agents, knowledge, tools, telephone numbers, channels, integrations, tickets, evaluation, and billing ) in plain language.
Core Chapters
| Section | Chapter | Description |
|---|---|---|
| 00. Concepts | Concepts & Glossary | Shared vocabulary and platform terminology |
| Mental Model | High-level system architecture and message lifecycle | |
| 01. AI Agents | Agents Overview | Agent builder foundations and the draft/publish/deploy lifecycle |
| Agent Types | Chat vs Voice agent capabilities and constraints | |
| Configuring Agents | Brand prompts, personalization, rules, and voice telephony settings | |
| Scenarios & Routing | Playbook definitions, triggers, and runtime intent routing | |
| Scenarios vs Knowledge Bases | Decision guide: when to use workflows vs searchable documents | |
| Versioning & Deployment | Managing version history, rollbacks, and active deployments | |
| Archiving Agents | Soft-retiring agents, tearing down voice assistants, and number detachment | |
| 02. Knowledge | Knowledge Bases Overview | Document ingestion, indexing states, and grounding attachments |
| 03. API Tools | API Tools Overview | Org-defined HTTP GET tools for live runtime lookups |
| 04. Telephony | Phone Numbers Overview | Telephony number inventory, provisioning, and agent assignment |
| 05. Integrations | Freshchat Integration | Connecting WhatsApp, Instagram, and Freshchat topics to agents |
| 06. Widgets | Widgets | First-party embeddable web chat widget styling and domain allowlists |
| 07. Support Desk | Tickets & Recordings Tags | Post-session summaries, audio recordings, and single/bulk tag management |
| 08. Evaluation | Evaluation Overview | Synthetic benchmarking and quality scoring test suites |
| 09. Spend & Usage | Billing & Usage Overview | Session telemetry, spend metrics, invoices, and payments |
🚀 Getting Started
Start with the Concepts & Glossary to get familiar with terms, then review the Mental Model before configuring your first AI Agent.