Operating ProcedureExecution layer

Assistable Multi-Channel Agent Deployment (Delivery)

A sequence with 8 steps: Create a sub-account for the client in the Assistable dashboard.

By InnovaAI ResearchPublished Updated

What are the steps?

sequence

Assistable Multi-Channel Agent Deployment (Delivery)

  1. 01

    Create a sub-account for the client in the Assistable dashboard

    Navigate to the 'Sub-accounts' section and click 'Create Sub-account'. Enter the client's business name and select the appropriate plan tier based on expected call volume.

  2. 02

    Configure the agent persona in the visual no-code builder

    In the agent builder, set the agent's name, voice, and tone. Use the persona settings to align with the client's brand voice. Save the persona before proceeding.

  3. 03

    Upload business FAQs and knowledge base documents

    In the 'Knowledge Base' tab, upload PDFs, text files, or URLs containing business information. The system will index these for query responses at $0.01 per query.

  4. 04

    Connect client's CRM and calendar integrations

    From the 'Integrations' menu, select native integrations like Salesforce, HubSpot, GoHighLevel, or Calendly. Authenticate with client credentials and map fields for lead capture and booking.

  5. 05

    Enable voice, SMS, WhatsApp, and chat channels

    In the 'Channels' settings, activate each channel. For voice, configure the phone number and set the per-minute rate (starting at $0.07/min). For SMS and WhatsApp, verify the numbers and set up message templates.

  6. 06

    Set up cross-channel memory and human handoff rules

    Enable the shared memory layer so conversations can continue across channels. Define handoff triggers (e.g., when the agent detects frustration) and route to a human agent via the dashboard or integrated helpdesk.

  7. 07

    Test the agent with simulated interactions

    Use the built-in testing tools to simulate a full conversation across channels. Verify that the agent correctly accesses the knowledge base, updates CRM records, and schedules appointments. Iterate on the flow as needed.

  8. 08

    Deploy the agent to production and monitor performance

    Publish the agent to go live. Monitor conversation logs and performance metrics in the dashboard. Set up alerts for high error rates or low satisfaction scores, and schedule monthly reviews with the client.