Warmly’s AI Studio makes it effortless for your team to build personalized AI chat agents that can engage visitors, answer questions, qualify buying intent, and even book meetings for you - 24/7.

Here's a quick guide to help you get your first Chat Agent up and running in just a few clicks!

🛠 Step-by-Step Instructions

Step 1: Create a new Chat Agent

This is where all your chat agents live.

1. Go to Inbound Agent → AI Studio in the left-hand sidebar.

2. Click Add Chat Agent to create a new agent.

This will open the configuration flow where you’ll build, train, and customize your agent.

Step 2: Upload Training Content

This is where your Chat Agent learns everything it needs to know about your product, your customers, and how you want it to communicate. The more relevant content you add, the smarter and more accurate your agent will be.

You can train your agent using a variety of content types, including:

• Product documentation – Teach your agent core features, functionality, and workflows.

• FAQs – Help the agent quickly answer common visitor questions.

• Sales playbooks – Let it follow your team’s qualification criteria, messaging, and objection-handling steps.

• Help center articles – Provide step-by-step guidance for support-related questions.

Your agent will use all of this information to respond with accurate, context-aware answers during conversations.

To begin, click the 'Add Content' button and choose the type of training content you want to upload.

Step 2a: Upload PDFs

You can upload internal or customer-facing PDFs directly into the AI Studio:

• Product sheets

• Onboarding guides

• Pitch decks

• Internal process docs

Once uploaded, the AI Studio automatically analyzes the content and adds it to your agent’s knowledge base.

Step 2b: Upload Website Links

Train your agent using live website content such as your:

• Marketing site

• Documentation site

• Blog posts

• Knowledge base

• Pricing pages

When adding a website link, you’ll be able to choose:

1. Train on Full Website or Single Page

• Full website: Ideal if your agent needs a broad understanding of your product and company.

• Single page: Best for highly focused or specialized training.

2. Exclude Certain Pages

If there are pages you don’t want your agent to learn from (e.g., outdated docs, blog posts, legal pages), you can manually exclude them. This keeps your agent trained only on the most accurate and relevant content.

Step 3: Configure Personality

Once your agent has been trained with the right content, the next step is shaping how it sounds and behaves. This is where you customize the agent’s personality, tone, and rules so it feels aligned with your brand and operates exactly the way you want.

Step 3a: Set Persona & Tone

Define the role, style, and personality of your Chat Agent. This determines how it communicates with visitors throughout a conversation.

You can specify traits such as:

• Professional and concise

• Friendly and conversational

• Playful and fun

• Empathetic and supportive

• Sales-driven and persuasive

• Technical and detail-oriented

You can also assign a role - for example:

• “Sales Development Representative”

• “Customer Support Specialist”

• “Product Expert”

• “Onboarding Assistant”

This helps your agent understand the context it should operate in and the type of language it should use.

Step 3b: Set Specific Rules & Constraints

Rules ensure your agent responds appropriately, stays within brand guidelines, and handles edge cases gracefully.

Use this section to establish:

1. Behavior Guidelines

• How formal or casual the agent should be

• How long messages should be

• Whether it should speak in first-person or third-person

• How to handle sensitive questions or off-topic conversations

2. Fallback Behavior

Define what the agent should do when it lacks the required information:

• Provide a generic helpful response

• Ask clarifying questions

• Redirect to a specific teammate

• Share a help center link

These rules act as the guardrails that keep your agent aligned with your expectations and ensure a high-quality experience for every visitor.

Step 4: Name your Chat Agent

The final step in setting up your agent’s configuration is giving it a name. Choose a name that reflects the agent’s role, tone, or purpose. A clear, descriptive name helps teammates instantly understand the agent’s responsibility.

Step 5: Add Chat Agent to your Chat workflow

To activate your agent in conversations:

1. Open the chat workflow where you want the agent to run.

2. Add the Start Infinite AI Chat step.

3. Select your newly created Chat Agent.

Step 6: Assign a goal to your Chat Agent

The chat agent should have a clear objective so it knows what success looks like in a conversation. Goals help the agent prioritize responses and navigate visitors toward the intended outcome.

Examples of goals include:

• Qualify leads

• Book meetings

• Answer product questions

• Route visitors to the right team

• Provide onboarding guidance

• Troubleshoot support issues

Once a goal is assigned, your agent will use its training data and personality settings to steer conversations toward completing that objective.

Step 7: Add Exit Conditions + Branching

Exit conditions tell the agent when to end the chat, move to a new workflow step, or trigger an action. Branching allows you to build dynamic conversation paths based on visitor responses.

Use this section to define:

Step 7a: Exit Conditions: When should the agent stop or transition?

Common exit conditions include:

• Lead fully qualified

• Visitor requested human assistance

• Visitor becomes unresponsive

• Agent cannot answer the question

Step 7b: Branching Logic

Create conversational paths that adapt to what the visitor says or does. For example:

• If the visitor expresses buying intent → qualify + offer to book a meeting

• If the visitor needs support → route to support flow

• If the visitor asks pricing questions → show pricing content

• If the agent detects confusion → offer clarifications or ask follow-up questions

Once you’ve built your branching flows, tested your updates, and published your changes, you’re all set! Watch your inbound conversations instantly become smarter, faster, and more efficient.