The AI Agent Studio provides a comprehensive suite of tools for organizations to create, test, deploy, and manage AI agents to automate customer service processes.

Freddy AI Agent now has the ability to think, reason, and take action to address customer inquiries from beginning to end. 

Within the AI Agent Studio, you will find:

  1. Skills Builder - A visual, no-code Skills Builder, enabling organizations to seamlessly create and deploy new skills for Freddy AI Agent, empowering it to perform autonomous tasks. These skills include order management, booking management, and subscription management, all of which necessitate connections with external systems, extensive workflow orchestration, and advanced thinking and reasoning capabilities.
  2. Skills Library- A library of pre-built skills by integrating popular third-party platforms like Shopify and Stripe, so that typical tasks like order tracking and returns processing can be quickly automated by the AI Agent. The following skills will be available at launch:
    • Shopify - Fetch Orders, Fetch Order Details, Cancel Order.
    • ShipStation - Fetch Orders.
    • ShipRocket - Fetch Orders, Track Orders.
    • Stripe - Cancel Subscription, Fetch Subscription Details.


This article contains

Define Skills to Match Business Workflows

A Skill is a reusable, no-code workflow that tells your AI Agent how to complete a specific multi-step process. Built with intuitive building blocks, Skills can:

  • Ask questions to collect data from users
  • Perform API calls to external systems
  • Branch logic using conditional paths
  • Personalize responses with variables
  • Escalate issues to human agents if needed

With Skills, your AI Agent becomes capable of executing actions, not just explaining them.

Trigger Skills Based on Customer Intent

Using the AI Agent Studio's Skills Builder, start a Skill automatically by defining natural language triggers. Just describe when the Skill should run.

Example: “Trigger when the customer wants to cancel their subscription.”

Freddy will monitor conversations for matching intents and route to the appropriate Skill at the right time.

Collect Information Dynamically

Use the Collect Info block to ask users relevant, context-aware questions and store their answers as variables.

Example: “Ask for the customer’s registered email address and subscription ID.”

These variables are then used throughout the Skill—to drive logic, personalize responses, or pass into an API call.

Trigger API Actions in Real-Time

With the API Action block, connect the AI Agent to your external tools—CRM, order systems, billing platforms, and more.

Admins can map local or global variables into the API URL, headers, or payload, and store returned values for downstream use.

Example: “Call the shipping provider’s API to retrieve order delivery status.”

All API actions are stored in a reusable API Action library—saving time and boosting consistency.

Branch Logic with Conditions

Add conditional logic using IF-THEN-ELSE statements to tailor workflows for different customer journeys.

Example: “If the plan is active, proceed with upgrade. If expired, offer renewal first.”

Use collected inputs or system-level attributes to control the flow and guide conversations intelligently.

Craft Personalized Responses

Send tailored replies using the Custom Response block. Choose between:

  • Scripted responses with static text and placeholders, or

  • AI-generated responses guided by instructions and variables.
    Example: “Your subscription for {{plan_name}} has been renewed and will expire on {{expiry_date}}.”

This flexibility helps ensure your bot always sounds helpful—never robotic.

Escalate to Human Agents When Needed

Insert an Agent Handoff block at any point to transition the conversation to a live agent.

Example: “If the refund policy is not applicable, route to a billing support rep.”

The Skill passes all context along, ensuring a seamless experience for both the agent and the customer.

Create Tasks with Natural Language (Coming Soon)

Use Tasks to auto-generate Skills with a single natural language instruction. Tasks combine blocks like Collect Info, API Action, and Custom Response into a single workflow—automatically.
Example: “Collect user name and email, check subscription status, and send renewal instructions.”

Tasks make it faster to set up simple Skills—but with some trade-offs in precision.

Reuse Skills and API Actions Across Agents

All created Skills and API Actions are stored in dedicated libraries and can be reused across multiple AI Agents. This promotes scale and consistency across your entire support operation.
Freshworks offers out-of-the-box (OOTB) Skills and APIs for common systems like Shopify, Stripe, and Shiprocket—so you can start automating immediately.

Apply Skills Across Real Use Cases

Here’s how organizations are already using Skills to power automation:

  • Automate Refund Requests: Collect order details, validate eligibility, trigger refund via API.

  • Manage Subscription Changes: Fetch current plan, offer upgrade/downgrade options, and confirm changes.

  • Streamline Customer Onboarding: Gather required details and sync to CRM automatically.

  • Support Payment Inquiries: Look up transaction status and respond with expected settlement dates.

Each Skill can be customized for tone, decision logic, and integrations, letting you match your exact business workflow.

Track Operational Impact

Skills drive tangible outcomes for your team:

  • Reduce resolution times by eliminating manual steps

  • Decrease ticket load by handling processes autonomously

  • Improve accuracy and personalization using real-time data

  • Accelerate time-to-market with no-code deployment

Admins can launch, iterate, and expand workflows without engineering dependencies.