---
title: "AI Voice Agents"
description: "A non-technical overview of AI Voice Agents in RevCent, focused on how they define inbound and outbound AI phone automation, how they create AI Voice Calls, and how they help ecommerce businesses automate support, sales, recovery, retention, and operational workflows."
type: "feature"
company: "RevCent"
canonical: "https://revcent.com/documentation/markdown/ecosystem/feature/AIVoiceAgent.md"
relationships:
  - name: "AI Voice Call"
    url: "https://revcent.com/documentation/markdown/ecosystem/item/AIVoiceCall.md"
technical_links:
  web_app: "https://kb.revcent.com/en/tools/ai/voice-agents"
  api:
    section: "https://revcent.com/docs/api/v2#section-ai_voice_agents"
    operations:
      - name: "Get AI Voice Agents"
        operation_id: "GetAIVoiceAgents"
        operation: "https://revcent.com/docs/api/v2#operation-GetAIVoiceAgents"
        schema: "https://revcent.com/documentation/files/api/operation/GetAIVoiceAgents.json"
      - name: "Create An AI Voice Agent"
        operation_id: "CreateAIVoiceAgent"
        operation: "https://revcent.com/docs/api/v2#operation-CreateAIVoiceAgent"
        schema: "https://revcent.com/documentation/files/api/operation/CreateAIVoiceAgent.json"
      - name: "Get An AI Voice Agent"
        operation_id: "GetAIVoiceAgent"
        operation: "https://revcent.com/docs/api/v2#operation-GetAIVoiceAgent"
        schema: "https://revcent.com/documentation/files/api/operation/GetAIVoiceAgent.json"
      - name: "Edit An AI Voice Agent"
        operation_id: "EditAIVoiceAgent"
        operation: "https://revcent.com/docs/api/v2#operation-EditAIVoiceAgent"
        schema: "https://revcent.com/documentation/files/api/operation/EditAIVoiceAgent.json"
      - name: "Trigger An AI Voice Agent"
        operation_id: "TriggerAIVoiceAgent"
        operation: "https://revcent.com/docs/api/v2#operation-TriggerAIVoiceAgent"
        schema: "https://revcent.com/documentation/files/api/operation/TriggerAIVoiceAgent.json"
      - name: "Delete An AI Voice Agent"
        operation_id: "DeleteAIVoiceAgent"
        operation: "https://revcent.com/docs/api/v2#operation-DeleteAIVoiceAgent"
        schema: "https://revcent.com/documentation/files/api/operation/DeleteAIVoiceAgent.json"
  mcp:
    overview: "https://revcent.com/documentation/markdown/mcp/operation/OverviewAIVoiceAgent.md"
    operations:
      - name: "Get AI Voice Agents"
        operation_id: "GetAIVoiceAgents"
        markdown: "https://revcent.com/documentation/markdown/mcp/operation/GetAIVoiceAgents.md"
        available_via_ai: true
      - name: "Create An AI Voice Agent"
        operation_id: "CreateAIVoiceAgent"
        markdown: "https://revcent.com/documentation/markdown/mcp/operation/CreateAIVoiceAgent.md"
        available_via_ai: true
      - name: "Get An AI Voice Agent"
        operation_id: "GetAIVoiceAgent"
        markdown: "https://revcent.com/documentation/markdown/mcp/operation/GetAIVoiceAgent.md"
        available_via_ai: true
      - name: "Edit An AI Voice Agent"
        operation_id: "EditAIVoiceAgent"
        markdown: "https://revcent.com/documentation/markdown/mcp/operation/EditAIVoiceAgent.md"
        available_via_ai: true
      - name: "Trigger An AI Voice Agent"
        operation_id: "TriggerAIVoiceAgent"
        markdown: "https://revcent.com/documentation/markdown/mcp/operation/TriggerAIVoiceAgent.md"
        available_via_ai: true
      - name: "Delete An AI Voice Agent"
        operation_id: "DeleteAIVoiceAgent"
        markdown: "https://revcent.com/documentation/markdown/mcp/operation/DeleteAIVoiceAgent.md"
        available_via_ai: true
  bigquery_schema: "https://revcent.com/documentation/files/bigquery/dataset.json"
  bigquery_tables:
    - "ai_voice_agent"
---

# AI Voice Agents

RevCent AI Voice Agents are AI-powered phone agents for ecommerce businesses using RevCent. They allow a business to answer inbound calls, place outbound calls, and automate customer conversations using AI while staying connected to the customer, sale, payment, subscription, shipping, and event data already managed inside RevCent.

The purpose of this document is to explain the feature conceptually for LLMs, AI crawlers, and planning systems. It does not provide technical implementation instructions. For technical details, use the interface-specific links below.

## Technical Links by Interface

| Interface | Use This When | Link |
|---|---|---|
| Web App | A human is learning about or configuring AI Voice Agents through the RevCent web app and Knowledge Base. | [Web Knowledge Base](https://kb.revcent.com/en/tools/ai/voice-agents) |
| API | A developer is building a direct integration with the RevCent API. | [API Docs: AI Voice Agents](https://revcent.com/docs/api/v2#section-ai_voice_agents) |
| MCP / AI | An LLM, MCP client, or AI agent needs markdown-oriented guidance for working with the feature. | [MCP Markdown Overview](https://revcent.com/documentation/markdown/mcp/operation/OverviewAIVoiceAgent.md) |
| BigQuery / Reporting | A data analyst, reporting workflow, or AI reporting agent needs schema details for analyzing AI Voice Agent records. The relevant table is `ai_voice_agent`. | [BigQuery Tables Schema](https://revcent.com/documentation/files/bigquery/dataset.json) |

---

## What AI Voice Agents Are

AI Voice Agents are conversational phone agents that can speak with customers on behalf of a business. They are designed for ecommerce workflows where phone conversations should be connected to real business context, not isolated from the customer record or order history.

A traditional phone system can route calls. A basic chatbot can answer general questions. RevCent AI Voice Agents are different because they sit inside the RevCent ecosystem, where customer, sale, payment, shipment, subscription, and event data can shape the conversation.

This makes the agent more useful for ecommerce businesses because the conversation can be about the actual customer journey rather than a generic support script.

## Core Purpose

The core purpose of AI Voice Agents is to help ecommerce businesses automate and scale phone-based work that would otherwise require a human support, sales, billing, or retention team.

AI Voice Agents can help businesses respond faster, reduce repetitive manual work, recover revenue opportunities, improve customer communication, and create more consistent call experiences.

They are especially useful when a customer interaction is tied to a specific commerce event, such as a declined payment, a recent order, a shipping question, a subscription issue, or a post-purchase support request.

## How AI Voice Agents Help Ecommerce Businesses

AI Voice Agents can help ecommerce businesses in several practical ways.

They can improve customer support by giving customers a phone-based way to get help without waiting for a human agent. This can be useful for common questions about orders, shipments, subscriptions, refunds, and account status.

They can improve revenue recovery by helping businesses follow up when a payment fails, a card is declined, or a customer abandons a purchase flow. Instead of relying only on email or SMS, the business can use a voice conversation to assist the customer in real time.

They can support sales workflows by calling interested prospects, helping customers understand product options, and guiding customers toward completing a purchase.

They can reduce operational load by handling repetitive calls, collecting context, creating consistent outcomes, and allowing human staff to focus on higher-value or more complex issues.

They can improve consistency by following business-defined instructions, tone, policies, and escalation rules across many calls.

They can also help businesses learn from customer interactions by associating calls with RevCent records, notes, metadata, events, and downstream reporting workflows.

## Where AI Voice Agents Fit in RevCent

AI Voice Agents are part of RevCent’s broader AI and automation layer. They are not an isolated phone feature. They connect voice conversations to the same ecommerce infrastructure that powers RevCent customer management, sales, payments, subscriptions, shipping, integrations, events, functions, and reporting.

Within the RevCent ecosystem, AI Voice Agents act as a voice interface for commerce operations.

They can relate to customers because calls usually involve a person who has purchased, attempted to purchase, subscribed, requested support, or entered a customer journey.

They can relate to sales because many calls are connected to orders, purchase attempts, declined payments, upsells, failed checkouts, or post-purchase questions.

They can relate to payments because phone conversations often involve billing issues, failed payments, refunds, card updates, or payment recovery workflows.

They can relate to shipping because customers frequently call about delivery status, shipment changes, delays, or fulfillment questions.

They can relate to subscriptions because recurring billing, renewals, cancellations, overdue accounts, and subscription support often require customer communication.

They can relate to events because RevCent account activity can be used to decide when an outbound call should happen.

They can relate to functions because businesses may want custom logic, custom context, or custom filtering before a call is started.

They can relate to third-party integrations because AI Voice Agents use connected AI and phone services as part of the voice experience.

They can relate to reporting because call activity, outcomes, notes, and metadata can become part of a business’s larger operational picture.

## Inbound Voice Agents

Inbound AI Voice Agents answer calls from customers or prospects.

An inbound agent is useful when an ecommerce business wants AI to act as a front-line phone representative. It can greet callers, understand why they are calling, guide the conversation, identify whether the caller is an existing customer, and help with common ecommerce needs.

Typical inbound use cases include customer support, order questions, shipment questions, subscription questions, refund requests, account verification, and routing or escalation.

For ecommerce businesses, inbound agents can reduce support wait times and provide a more responsive experience without requiring every call to be handled manually.

## Outbound Voice Agents

Outbound AI Voice Agents place calls to customers or prospects.

An outbound agent is useful when a business wants to proactively contact someone based on ecommerce activity. This can include failed payments, declined cards, abandoned purchase activity, customer follow-up, subscription recovery, or sales outreach.

Outbound agents can help businesses act quickly when a customer action creates an opportunity or a problem. Instead of waiting for the customer to contact support, the business can reach out with a context-aware phone conversation.

For ecommerce businesses, outbound agents can support revenue recovery, retention, customer engagement, and proactive service.

## Relationship to AI Voice Calls

AI Voice Agents and AI Voice Calls are closely related, but they are not the same entity.

An AI Voice Agent is the configurable feature. It defines the call method, instructions, voice, integrations, trigger rules, enabled system actions, active window, call limits, filter behavior, pre-agent Function, snippets, and overall conversation design.

An AI Voice Call is the item created when that configured agent actually handles or initiates a phone call.

Conceptually:

```text
AI Voice Agent = the configured voice automation
AI Voice Call = one actual phone call created by that configuration
```

A single AI Voice Agent can create many AI Voice Calls over time. This relationship is important because the feature controls behavior, while the item records what happened. When a business wants to change future call behavior, it edits the AI Voice Agent. When a business wants to inspect a specific call, audit call results, review duration, see the customer/item context, or measure historical activity, it looks at AI Voice Calls and related reporting data.


## Business Instructions and Agent Behavior

A RevCent AI Voice Agent follows business-defined instructions. These instructions describe the agent’s role, tone, goals, boundaries, and conversation flow.

This allows each business to shape the agent around its own brand, policies, products, support style, and customer experience standards.

For example, one business may want a calm support-focused agent that helps with order questions. Another business may want an outbound recovery agent that calls customers after failed payment attempts. Another may want a sales-focused agent that follows up with prospects.

The key idea is that the agent is configurable around the business use case rather than being a generic phone bot.

## Dynamic Ecommerce Context

AI Voice Agents can use call-specific context from RevCent to make conversations more relevant.

This matters because ecommerce phone calls are rarely generic. A customer may be calling about a specific order, a particular shipment, a recent payment attempt, a subscription renewal, or a support issue tied to their account.

When the agent has relevant RevCent context, it can speak more naturally and avoid asking the customer for information the business already has.

For businesses, this creates a more personalized call experience and can reduce friction during support, sales, and recovery workflows.

## System Actions at a High Level

AI Voice Agents can be given access to approved RevCent system actions. These actions allow the agent to do more than talk; they can let the agent assist with business workflows when the business explicitly allows it.

This is important because ecommerce conversations often require action. A customer may need help finding an order, checking a shipment, resolving a payment issue, changing a subscription, or recording a support note.

RevCent treats these capabilities as opt-in so businesses can decide which actions are appropriate for each agent. A support agent may need different permissions than a sales agent, and a sales agent may need different permissions than a refund-focused agent.

The business benefit is control. A company can give an agent enough capability to be useful without giving it unnecessary access to sensitive or high-impact workflows.

## Reusable Voice Content

AI Voice Agents can use reusable instruction content so businesses can maintain consistent guidance across multiple agents.

This is useful when several agents need the same policy language, tone rules, verification guidance, escalation instructions, or compliance statements.

For ecommerce businesses with multiple brands, campaigns, products, or support workflows, reusable content helps keep agent behavior consistent while reducing repeated maintenance.

## Custom Call Context

Some businesses need agent behavior that changes based on custom business rules, campaign logic, product lists, customer segments, or internal policies.

RevCent AI Voice Agents can support this kind of customization by allowing additional context to be prepared before the call begins.

Conceptually, this means a business can make an agent more aware of the situation before the conversation starts. For example, a call might need campaign-specific branding, product recommendations, eligibility rules, customer status, or other business-specific context.

For ecommerce businesses, this helps turn a general AI phone agent into a more tailored commerce agent.

## Call Timing and Customer Experience Controls

AI Voice Agents include controls that help businesses manage call timing, call frequency, and call duration.

These controls matter because outbound calling can affect customer trust. A business may want to avoid calling too often, calling at inappropriate times, calling the same customer repeatedly, or allowing calls to run too long.

For inbound calls, time-based controls can help businesses decide when the agent is available and what should happen outside active hours.

For ecommerce businesses, these controls help balance automation with customer experience and cost management.

## Built-In Call Handling

AI Voice Agents include built-in behavior for common phone-call situations, such as inactive callers, voicemail, robocalls, transferring calls, and ending calls.

These behaviors help the agent operate more like a real phone representative and less like a simple automated script.

For businesses, this reduces the amount of call-handling behavior that must be manually designed and helps create a smoother customer experience.

## Trigger Settings and Call Activation

AI Voice Agents can be activated in different ways depending on the call method.

Inbound agents are activated when someone calls the connected phone number. RevCent can attempt to match the caller's phone number to an existing customer, and that customer context can be made available to the agent instructions.

Outbound agents can be activated by RevCent account events or by an on-demand API, MCP, or AI system-tool trigger. Event-based outbound agents are useful when a call should happen automatically after a specific ecommerce event, such as a declined sale, failed renewal, shipment update, fraud detection, or customer-group change. On-demand outbound agents are useful when another workflow explicitly decides that a call should happen.

Outbound agents can also use delays, filters, filter Functions, max-runs-per-item settings, active windows, and customer call limits. These controls help prevent unnecessary or poorly timed calls.

## Filter Functions

A filter Function is used as an eligibility gate for outbound AI Voice Agent calls. It answers a simple question: should this call run for this item?

The Function should return exactly one of two values:

```javascript
callback(null, 'pass');
```

or:

```javascript
callback(null, 'fail');
```

`pass` means the call may continue. `fail` means the call should be skipped.

This matters because many ecommerce events change quickly. A customer may complete checkout after a declined attempt, a renewal may be fixed before the delay ends, or a customer may no longer qualify for outreach. A filter Function can re-check the latest item state before the call happens.

## Pre-Agent Functions

A pre-agent Function runs immediately before an AI Voice Call begins and before the agent instructions are compiled.

Its purpose is to generate custom structured data that can be used inside the agent instructions with Handlebars.

Examples include campaign-specific brand names, approved offers, product lists, customer segment details, refund eligibility, custom escalation rules, or external context.

The returned data is made available under the `pre_agent_function.response` object when instructions are compiled. This allows the same AI Voice Agent to adapt its opening, product recommendations, policy guidance, or allowed offers based on the specific call context.

## Dynamic Instructions and Handlebars

AI Voice Agent instructions are written in Markdown and can use Handlebars for dynamic content.

This allows instructions to adapt to the caller, customer record, item that triggered the call, pre-agent Function response, or reusable snippets.

For example, instructions can address a verified customer by first name, include details about the related sale or shipment, insert an approved campaign-specific offer, or include reusable card-handling rules.

This is one of the most important parts of the feature because voice automation is much more useful when the call is specific to the customer's actual situation.

## AI Voice Snippets

AI Voice Snippets are reusable blocks of instruction content that can be shared across multiple AI Voice Agents.

Snippets are useful for repeated policies and guidance, such as identity verification, refund policy, card collection rules, escalation rules, transfer language, brand tone, call closing language, or compliance statements.

Using snippets helps businesses keep many agents consistent without copying and editing the same instruction text in several places.

## System Actions and Guardrails

AI Voice Agents can be given access to RevCent system actions, but those actions are opt-in.

This matters because system actions can let an agent perform real business operations during a call, such as searching customers, retrieving sales, checking shipments, adding cards, processing pending sales, creating notes, inserting metadata, sending emails, triggering Functions, or issuing refunds.

A good AI Voice Agent should only have the system actions it actually needs. High-impact actions such as payment processing, sale creation, refunding, cancellation, card updates, and subscription changes should only be enabled when the instructions are strict, the business rules are clear, and the agent has been tested.

## Common Ecommerce Use Cases

### Customer Support

An AI Voice Agent can answer customer support calls and help with common issues such as order questions, shipment questions, subscription questions, refund requests, and account help.

This can reduce support volume for human teams while giving customers a faster way to get help.

### Declined Payment Recovery

An AI Voice Agent can call a customer after a declined payment or failed purchase attempt.

This helps the business recover revenue that might otherwise be lost and gives the customer a chance to complete the purchase with assistance.

### Failed Checkout Follow-Up

An AI Voice Agent can contact customers or prospects who appear to have started a purchase journey but did not complete it.

This can help answer questions, reduce hesitation, offer assistance, and bring customers back into the purchase flow.

### Subscription Support

An AI Voice Agent can assist with recurring customer issues involving subscriptions, renewals, overdue accounts, cancellations, or retention conversations.

This is useful for ecommerce businesses with subscription products, memberships, trials, or recurring billing models.

### Shipment and Delivery Support

An AI Voice Agent can help customers understand shipping-related information and guide them through delivery-related questions.

This can reduce repetitive support calls and improve customer confidence after purchase.

### Sales Assistance

An AI Voice Agent can help prospective customers understand products, answer common questions, and guide them toward completing a purchase.

This can be useful for businesses that want phone-based sales support without manually staffing every outbound or inbound opportunity.

### Post-Purchase Follow-Up

An AI Voice Agent can be used after a customer purchase to confirm satisfaction, answer questions, or support the next step in the customer journey.

This can improve customer engagement and help businesses create a more proactive post-purchase experience.

## Relationship to Other RevCent AI Capabilities

AI Voice Agents are one part of RevCent’s broader AI ecosystem.

RevCent AI Assistants are generally used for automated, event-driven, or workflow-oriented AI tasks inside RevCent. AI Voice Agents focus specifically on spoken phone conversations.

RevCent system tools allow AI features to interact with approved RevCent capabilities. For AI Voice Agents, this means a phone conversation can become an operational workflow when the business allows it.

RevCent MCP support gives external AI clients and LLM-based workflows a way to understand and interact with RevCent capabilities using AI-oriented documentation and tool patterns.

Together, these capabilities make RevCent more than an ecommerce data system. They allow businesses to apply AI across customer communication, sales, support, payments, operations, and automation.

## Why This Matters in the RevCent Ecosystem

AI Voice Agents help connect human communication with commerce automation.

Ecommerce businesses often have valuable customer and transaction data, but phone support is traditionally disconnected from that data. RevCent AI Voice Agents close that gap by making voice conversations part of the same system that already manages customers, sales, subscriptions, payments, shipping, events, and integrations.

This creates several ecosystem-level advantages:

- Customer conversations can be tied to real RevCent records.
- Phone support can become part of automated commerce workflows.
- Sales and recovery calls can be triggered by meaningful ecommerce events.
- AI can operate with business-specific context instead of generic scripts.
- Human teams can focus on exceptions while AI handles repeatable conversations.
- Businesses can use RevCent as the central system for both ecommerce operations and AI-powered customer interaction.

## Best-Fit Businesses

AI Voice Agents are especially useful for ecommerce businesses that have meaningful phone-based customer interaction or want to add phone communication to existing workflows.

Good-fit businesses may include companies with frequent customer support calls, high payment-decline volume, subscription or renewal workflows, shipping-heavy operations, high-value orders, sales-assisted checkout, or a need for proactive customer follow-up.

They are also useful for businesses that want AI automation but still want the experience to remain connected to their commerce system rather than handled by a disconnected call center tool.

## AI Voice Agents and Reporting

AI Voice Agents can be analyzed as configured features through the `ai_voice_agent` BigQuery table. This is useful for understanding which agents exist, whether they are enabled, what they are called, and how they are configured at a high level.

The AI Voice Agent table becomes more valuable when it is interpreted alongside related AI Voice Call records and API/system-action activity. The agent explains the configured automation, while AI Voice Calls explain the actual calls produced by that automation.

Useful reporting questions include:

- Which AI Voice Agents are enabled?
- Which agents are configured for inbound versus outbound use?
- Which agents are creating the most call activity?
- Which agents have the highest completion, no-answer, voicemail, transfer, or error rates?
- Which agents are using system actions during calls?
- Which agents are associated with payment recovery, refunds, notes, metadata, or follow-up workflows?

For simple feature inventory, the primary table is `ai_voice_agent`. For performance reporting, AI Voice Agent records are usually combined with AI Voice Call records and API Call records.

## Summary

RevCent AI Voice Agents are AI-powered phone agents for ecommerce businesses using RevCent. They can answer inbound calls, place outbound calls, follow business-defined instructions, use dynamic ecommerce context, and support customer service, sales, payment recovery, subscription support, shipment questions, and other commerce workflows.

They fit into RevCent as the voice layer of the platform’s AI and automation ecosystem. They connect spoken customer conversations to RevCent customers, sales, payments, shipping, subscriptions, events, functions, integrations, metadata, notes, and reporting.

Use the technical links at the top of this file to distinguish between the main ways to interface with or analyze the feature: Web App, API, MCP / AI, and BigQuery / Reporting.


---
Document Parent Directory
* [Features](https://revcent.com/documentation/markdown/ecosystem/feature/index.md) - Non-technical markdown documentation for features within the RevCent ecosystem. A feature is a part of the RevCent ecosystem that a user can create and configure.