AUP block notification emails will now be sent to:

• The Agency Owner

OR

• Admins associated with the specific Sub-account

Fallback Logic

If no Agency Owner or Sub-account Admin is available, the system will send the email randomly as a fallback mechanism.

What's New

The AI Agent action now supports Update Custom Value as a built-in tool. The agent can evaluate context during execution and write directly to custom values inside a sub-account — replacing complex If/Else trees that previously required dozens of branches and actions.

• Eliminates the need for large If/Else trees built solely to set custom values based on conditions like day of the week, contact data, date formatting, timezone, or pipeline stage.
• Supports the familiar "Let AI decide" vs. static value pattern already used in other AI Agent tools.

How to Use

1. Select Update Custom Value as a tool inside the AI Agent action.
2. Custom Value: Pick a specific custom value to update, or set it to "Let AI decide" so the agent chooses the right field based on its reasoning.
3. New Value: Enter a fixed value or variable reference, or set it to "Let AI decide" so the agent generates the value based on your prompt instructions.

Common configurations:

• Custom Value static + Value AI-decided: Update a specific field with whatever the agent determines. This is the most common setup.
• Both AI-decided: The agent picks which custom value to update and what to set it to — suited for multi-field use cases where the target field depends on context.

Why This Matters

Workflows that previously needed branching logic for every possible condition — day-based schedules, formatted dates, rolling promo codes, context-aware values — now resolve in a single agent step. Less maintenance, fewer actions, faster execution, and lower complexity across your automations.

What's New

The AI Agent action now supports Knowledge Base Search as a built-in tool. During execution, the agent can search your native knowledge base in real time — pulling in relevant business context to improve decision-making and response accuracy.

• Connects directly to your native knowledge base, where your business context already lives: service details, pricing, offerings, booking rules, service menus, availability, FAQs, and more.
• Retrieves only the relevant information when needed, instead of loading the full system prompt with static content.
• Replaces the common workaround of pasting entire FAQ lists and business details into the prompt — which inflated the context window and produced less accurate results.

How to Use

1. Select Knowledge Base Search as a tool inside the AI Agent action.
2. Choose which knowledge bases the agent should search.
3. Define a static search query, or let the AI decide the query dynamically based on the conversation.
4. Set the number of chunks the agent retrieves from the knowledge base.

Why This Matters

Rather than front-loading the prompt with pages of business context, the agent now fetches exactly what it needs at runtime. This keeps the context window lean, reduces token usage, and delivers more accurate, grounded responses. Business knowledge stays in one place — the knowledge base — and the agent taps into it on demand.

What's New

The AI Agent action now supports Knowledge Base Search as a built-in tool. During execution, the agent can search your native knowledge base in real time — pulling in relevant business context to improve decision-making and response accuracy.

• Connects directly to your native knowledge base, where your business context already lives: service details, pricing, offerings, booking rules, service menus, availability, FAQs, and more.
• Retrieves only the relevant information when needed, instead of loading the full system prompt with static content.
• Replaces the common workaround of pasting entire FAQ lists and business details into the prompt — which inflated the context window and produced less accurate results.

How to Use

1. Select Knowledge Base Search as a tool inside the AI Agent action.
2. Choose which knowledge bases the agent should search.
3. Define a static search query, or let the AI decide the query dynamically based on the conversation.
4. Set the number of chunks the agent retrieves from the knowledge base.

Why This Matters

Rather than front-loading the prompt with pages of business context, the agent now fetches exactly what it needs at runtime. This keeps the context window lean, reduces token usage, and delivers more accurate, grounded responses. Business knowledge stays in one place — the knowledge base — and the agent taps into it on demand.

What's New

The AI Agent action now supports Knowledge Base Search as a built-in tool. During execution, the agent can search your native knowledge base in real time — pulling in relevant business context to improve decision-making and response accuracy.

• Connects directly to your native knowledge base, where your business context already lives: service details, pricing, offerings, booking rules, service menus, availability, FAQs, and more.
• Retrieves only the relevant information when needed, instead of loading the full system prompt with static content.
• Replaces the common workaround of pasting entire FAQ lists and business details into the prompt — which inflated the context window and produced less accurate results.

How to Use

1. Select Knowledge Base Search as a tool inside the AI Agent action.
2. Choose which knowledge bases the agent should search.
3. Define a static search query, or let the AI decide the query dynamically based on the conversation.
4. Set the number of chunks the agent retrieves from the knowledge base.

Why This Matters

Rather than front-loading the prompt with pages of business context, the agent now fetches exactly what it needs at runtime. This keeps the context window lean, reduces token usage, and delivers more accurate, grounded responses. Business knowledge stays in one place — the knowledge base — and the agent taps into it on demand.