Vatel AI Documentation

website·Indexed 32 days ago

https://docs.vatel.ai

REST and WebSocket APIs for building and managing voice agents. Includes tools for agent creation, session token generation, real-time voice sessions, SIP trunking, Twilio integration, and an Agent Editor for designing conversation and logic flows.

Tokens
29.4K
Snippets
119
Records
241
Agent score
57%

What's inside Vatel AI

  1. Overview of Logic Flows

    A logic flow is a graph of nodes designed to perform a specific task (e.g., looking up data, calling an API, or running branching logic). Logic flows act as tools that an agent invokes from the conversation flow to execute actions. Every logic flow has exactly one Start node as its entry point and ends when execution reaches a node with no outgoing connection. Variables set inside a flow are available call-wide and can be referenced by the conversation flow after the logic flow completes.
  2. Overview of Agent Editor Graph Node Types

    Nodes are the building blocks of the Agent Editor graph and are divided into two main categories:

    1. Convo (Conversation Flow): Defines what the agent says and does during a live call.
    2. Logic Flow: Handles branching, data manipulation, integrations, and internal processing steps.

    Key Constraints:

    • Function nodes can be used in both Convo and Logic flows.
    • Flow branching nodes (Condition, Categorize, Merge, Flow) are exclusive to Logic flows.
    • Edges and Conditions: All nodes support an optional when field, which is a plain-language condition (e.g., "caller has confirmed their email") that the agent evaluates before entering the node.
  3. Overview of Vatel AI Platform Capabilities

    Vatel AI is a platform for creating voice agents capable of handling natural dialogue, ranging from simple inquiries to complex multi-step workflows. The platform provides a full lifecycle for voice agents: designing and configuring prompts and flows, connecting agents to telephony and messaging channels, and monitoring performance via call analytics and transcripts.
  4. Deploy Vatel AI agents to production

    Vatel AI agents can be deployed into production environments using two primary methods: telephony integration for voice capabilities and SDKs for application-level integration. For telephony, users can connect via Twilio or SIP. For application integration, SDKs are available for Python, JavaScript (Node.js or browser), and Go.
  5. Integrate Vatel AI with n8n for Workflow Automation

    Use the n8n Public API integration to connect Vatel AI agents with n8n for complex workflow automation and third-party service integrations.
  6. Understand the types of tools available for Vatel agents

    Tools extend an agent's capabilities beyond basic conversation flow, allowing for search, mathematics, custom logic, and API-backed actions. Vatel provides three distinct types of tools:

    1. Function Nodes: Created within the agent editor. Any function node added to a conversation graph is automatically converted into a tool available to the agent.
    2. System Tools: Pre-built tools provided by the platform (e.g., web search, math operations, current time) that can be activated on a per-agent basis.
    3. Client-defined Tools: Custom tools defined and exposed by the developer. These are only accessible via the API and are used when the developer controls the client or the integration.
  7. Understand the Agent Editor graph-based architecture

    The Agent Editor uses a graph of nodes to define an AI agent's behavior. This architecture allows the agent to move through a directed graph during a call, branching or looping based on user input rather than following a rigid script. The graph consists of nodes that structure the logic and determine what the agent says and does.
  8. Integrate Vatel AI with HubSpot Automation Sequences

    Connect Vatel AI agents to HubSpot to manage sequences and email automation via the HubSpot Automation Sequences API.
  9. Integrate Twilio Voice and SMS with Vatel AI

    The Twilio integration allows Vatel AI agents to handle voice calls and SMS. Billing for usage (calls, SMS, and phone numbers) is handled directly by Twilio. To set up the integration, you need a Twilio account and a Vatel AI organization with at least one agent.