How to Set Up the Support Agent in FlexiAgent

The Support Agent is a pre-built AI agent designed to handle your customer support conversations — answering questions, resolving common issues, and helping customers instantly, at any hour. Once configured with your business information, it can reply to most customer queries within seconds.

Think of the Support Agent as a support rep who never sleeps and never has a queue. A human support team answers one ticket at a time, during working hours. The Support Agent reads your business knowledge once and then answers every customer, immediately, all at once — even at 3 AM on a Sunday.

Why Use the Support Agent?

Here's the kind of impact this agent can have, based on real usage: inside FlexiFunnels itself, more than 5,000 chats every week are resolved through AI agents in seconds. Around 80% to 90% of customer queries can be replied to within seconds — often even more.

For your business, that means:

• Instant answers — customers get responses in seconds instead of waiting hours for a reply
• Better user experience — quick, accurate help keeps customers happy
• Lower refund rates — many refunds happen because customers get stuck and can't get help in time
• Less manual workload — your team only handles the queries the agent can't
• 24/7 coverage — support keeps running even when your team is offline

Your chat volume may be higher or lower depending on how many customers you have and what kind of queries you receive — but whatever the volume, the agent handles the repetitive questions so your team doesn't have to.

Part 1: Before You Start — Two Simple Setup Steps

Step 1: Create Your Product Knowledge Base File

Prepare a knowledge base file with the information your customers ask about most, such as:

• Products or services and their features
• Step-by-step help for common issues
• FAQs your support team answers repeatedly
• Business policies (refunds, cancellations, guarantees)
• Pricing and plan details
• Contact information and escalation options

Save it as a Markdown (.md) file — the only format the Vault currently supports. For example:

# Power House Gym — Support Knowledge Base

## Common Issues
Q: I can't access the gym after 10 PM.
A: 24/7 access is included in Premium and Elite plans only.
   Basic members have access from 6 AM to 10 PM.

## Refund Policy
Memberships can be cancelled with 30 days notice.
Refunds are processed within 5-7 business days.

## Contact
Phone: (555) 123-4567 | Email: support@powerhousegym.com

? Shortcut: Let AI build your knowledge base with our ready-made prompts

You don't have to write your knowledge base file yourself. We've created Vault Knowledge Builder prompts — one for the Sales Agent and one for the Support Agent — that turn your AI assistant into an interviewer that writes the file for you:

Open the Vault Knowledge Builder for Sales & Support Agents 

How to use it:

1. Open the document above — it contains both prompts (Sales Agent and Support Agent).
2. Copy the full prompt for the agent you're setting up. Copy everything — don't skip any part of it.
3. Open Claude, ChatGPT, or Gemini — whichever AI you use — paste the prompt into a new chat and send it.
4. Answer the AI's questions — share your website URL, a help desk link, or simply type your business details. It will keep asking until it has everything it needs.
5. The AI generates your knowledge base file(s) in Markdown. The Sales prompt typically creates one .md file (or one per product if you sell multiple products). Download them all and upload to the Vault (Step 2 below).

Tip: The more detailed your answers — real prices, real FAQs, real policies — the better your agent's responses will be. Always review the generated file(s) before uploading. For the full walkthrough, see "Create Vault Files Using Our Given Prompts" in Related Articles.

Step 2: Upload the Knowledge Base File to the Vault

After creating your knowledge base file:

1. Open the Vault from the left sidebar (under Data).
2. Click the + (Add) button in the top-right corner and upload your file.
3. Associate it with your Support Agent — in the upload window, choose Agent as the destination and select your Support Agent (for example, Sid-Support).

Once uploaded, the Support Agent will use this information to answer customer questions accurately.

Part 2: Start Chatting with the Support Agent

That's the setup done. Now you're ready to use the agent.

Step 1: Open the Chat Tab

Navigate to the Chat tab from the left sidebar of your FlexiAgent account.

Step 2: Select the Support Agent

Click Select agent at the top and choose your Support Agent from the dropdown of available agents.

Step 3: Start a New Chat

Click + New Chat if you're not already on a new conversation, and you can simply start chatting with it. That's it — that's all you need to do.

Step 4: Test the Agent

Ask questions the way your customers would. For example:

• How do I get started with your product?
• I'm having trouble logging in — what should I do?
• What is your refund policy?
• Do you offer support after purchase?
• How do I contact a human if I need more help?

The Support Agent will respond using the information stored in your connected knowledge base.

Example Use Case

A customer gets stuck at midnight and asks:

"I can't log in to my account — what should I do?"

Instead of waiting until your team comes online in the morning, the Support Agent instantly retrieves the troubleshooting steps from your knowledge base and walks the customer through the fix. The customer is unstuck in seconds — which is exactly how faster answers lead to happier customers and fewer refunds.

Best Practices

• Build the knowledge base from real tickets. Your support history is a ready-made list of what customers actually ask.
• Keep the knowledge base updated. New feature? New policy? Update the file and the agent starts using it immediately.
• Test the agent after every major update — ask it a few real customer questions and verify the answers.
• Review conversations regularly in the Sessions tab to spot questions the agent couldn't answer well — then add those answers to your knowledge base.
• Include an escalation path (email or phone) in your knowledge base so the agent can point customers to a human when needed.

Common Situations & Quick Fixes