Using ilert AI Voice Agent

The AI voice agent in ilert enables human-like phone interactions for faster, smarter incident handling. It understands caller intent, gathers key information, and provides real-time updates using connected service data. This helps reduce manual triage, streamline incident reporting, and improve MTTR.

Add the AI voice agent to your call flow

1. Navigate to Call routing and open or create a call flow.

2. Click the + button to add a new node.

3. From the node list, select AI voice agent.

4. The AI voice agent configuration panel will open on the right.

Configure the greeting and agent name

1. In the Optional title field, enter a name to label the node (e.g., “AI Virtual Assistant”).

2. In the Greeting field, write a message to welcome the caller (e.g., “Welcome to ACME, this is Alex, your virtual assistant. How may I assist you today?”).

Define caller intents

1. Under Intent configuration, click + Add intent.

2. Select from existing intents like:

   • Report critical incident

   • Report system outage

   • Report security breach

   • Request technical support

   • General inquire

3. Or click Create new to define a custom intent with a description.

4. Click Add to save.

Each intent becomes a branch in your call flow for further configuration.

Configure information gathering

1. In the Information gathering section, select the data the agent should collect:

   • Caller name

   • Contact number

   • Email address

   • Incident description

   • Affected services

2. To add a custom field:

   • Click + Add custom field

   • Enter a label (e.g., Customer ID)

   • Choose the field type (e.g., Text)

   • Optionally, make it required and add a sample question

   • Click Add

Enable knowledge enrichment (optional)

1. Toggle on Enable knowledge enrichment.

2. Choose which data types to include:

   • Open incidents

   • Maintenance windows (scheduled and ongoing)

   • Service status (even if there are no incidents)

3. Select data sources (e.g., API, UI, AI, Payment).

This allows the AI to respond with dynamic updates based on real-time incident data.

Route caller based on intent

1. After adding intents, each intent appears as a separate branch in the call flow.

2. Click the + icon under each intent to define follow-up actions, such as:

   • Route call

   • Create alert

   • Play an audio message

   • Ask for PIN

   • Block numbers

3. Configure a Fallback action for unmatched inputs.

4. Click Save to apply changes to your call flow.

Build the complete call flow

Once the AI voice agent is configured, you can create the full call handling flow:

1. Incoming call node – Entry point for all calls.

2. AI voice agent node – Greets caller, identifies intent, and collects requested information.

3. Intent branches – Each intent appears as its own branch. Example:

   1. Report critical incident – Route to a live agent or create a high-priority alert.

   2. Report system outage – Trigger a recorded status update and log the incident.

   3. Fallback – Handle unknown inputs with a standard message or voicemail option.

4. Actions under each branch – Click the + icon under each intent to define what happens next, such as:

   • Route call

   • Create alert

   • Play audio message

   • Ask for PIN

   • Block numbers

5. End call node – Close the call after actions are complete.

This is how a Call Flow could look:

FAQ

Can I add multiple intents to one AI voice agent? Yes, you can add and configure multiple intents per agent, each with its own branch and follow-up logic.

Is knowledge enrichment required? No, but enabling it helps the AI respond to callers with real-time service data.

What languages does the AI voice agent support? The AI voice agent supports English and German. Additional language support may be added in future releases.

Can I test the AI voice agent before going live? Yes, you can preview and simulate your call flow directly within ilert before assigning a real number.