Rafah is a browser extension that seamlessly injects state-of-the-art AI features into customer engagement platforms. Say goodbye to the hassle of switching between tools and embrace the future with Rafah's in-platform AI capabilities.
-
🛡️ Stay on Platform: Access sophisticated AI tools without ever leaving your customer engagement platform.
-
🔓 Break Through Closed Systems: Even the most resistant platforms can now enjoy external functionality.
-
⚙️ Effortless Configuration: Get started with AI in just a few clicks.
-
🤖 AI Chatbots: Enhance customer interaction with intelligent, conversational chatbots.
-
💬 Automated Responses: Deliver instant replies to common inquiries with AI-generated responses.
- Google Chrome
- Microsoft Edge
-
Check the Getting Started documentation for Vectara to start a new corpus.
-
You can use the demo data here
-
Create a new Vectara API Access
-
Register an account in Anyscale and get your API key.
-
Download the Rafah project zip file. Extract it to a folder.
-
Open the Chrome extensions (
chrome://extensions) page. (Run with Edge browsers too) -
Turn on the developers mode.
- Click on load unpacked button, and select the
src/chromefolder from the extracted zip file.
- Now you can see the Rafah extension in the extensions list.
- Pin the extension to the toolbar for easy access.
- Open the Rafah options page through the extension icon.
- Fill the configuration form with the required details, and click on the
Savebutton.
If you did not setup yet a customer engagement platform, you can use the following demo to test Rafah.
-
Open the Chatwoot demo page.
-
Enter the login credentials
Email: [email protected]
Password: ##Agent-user-1##
- Create an empty HTML page and inject the code below.
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>DEMO</title>
</head>
<body>
<script>
(function(d,t) {
var BASE_URL="http://18.236.158.230:3000";
var g=d.createElement(t),s=d.getElementsByTagName(t)[0];
g.src=BASE_URL+"/packs/js/sdk.js";
g.defer = true;
g.async = true;
s.parentNode.insertBefore(g,s);
g.onload=function(){
window.chatwootSDK.run({
websiteToken: 'JArqz7KPJrEG7Ky1hrMdcD1r',
baseUrl: BASE_URL
})
}
})(document,"script");
</script>
</body>
</html>-
You will need in the extension options page to set the
Platform URLvalue tohttp://18.236.158.230:3000 -
Open the HTML page in the browser, and start chatting with the bot.
-
You can track the new messages from the inbox.
Open any inbox messages for a customer, and you find new augmented elements into the web page.
Next image is demonstrating some of the augmented elements.
-
Each new messages will be classified automatically to a topic name.
-
You can ask to summerize the whole conversation with the customer. The summerization will include a section for unresponded messages for the customer.
-
You can switch the AI composer to generate a response for the customer. It can be used as a suggestion for the agent to respond to the customer.
-
If the user's message is requiring external data resources from Vectara, you can click on this icon to get the search results.
-
The composed message can be editable by the agent before sending it to the customer.
Community contributions make Rafah better for everyone. To contribute:
- Fork the project repository.
- Make your changes.
- Create a pull request.
Rafah is made available under the Apache License 2.0
Kudos to the contributors, users, and supporters of Rafah who help enhance the dynamics of customer engagement platforms with AI.
This project was a part of Vectara Hackathon for developing RAG applications.