https://github.com/telexintegrations/support-ai

Leveraging Retrieval-Augmented Generation (RAG) and Google Gemini as the Large Language Model (LLM) to develop a Support AI that processes documents and text inputs from companies, enabling it to provide accurate, context-aware responses to customer inquiries.
https://github.com/telexintegrations/support-ai

Last synced: about 1 year ago
JSON representation

Leveraging Retrieval-Augmented Generation (RAG) and Google Gemini as the Large Language Model (LLM) to develop a Support AI that processes documents and text inputs from companies, enabling it to provide accurate, context-aware responses to customer inquiries.

• Host: GitHub
• URL: https://github.com/telexintegrations/support-ai
• Owner: telexintegrations
• Created: 2025-03-10T10:04:55.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-03-27T17:41:33.000Z (over 1 year ago)
• Last Synced: 2025-03-27T18:30:27.520Z (over 1 year ago)
• Language: Go
• Size: 2.53 MB
• Stars: 2
• Watchers: 2
• Forks: 1
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Telex Support AI

## Overview

Telex Support AI brings intelligent, real-time support assistance to your Telex organization. Powered by AI, it enhances customer support by leveraging a dynamic knowledge base consisting of company documents and support history. This integration acts as an AI-driven assistant within Telex, providing instant responses to queries and improving overall efficiency.

## Key Features

### 🚀 Instant AI-Powered Responses

- Get immediate, context-aware answers directly within your dedicated Telex channel.

- Reduce response time by automating repetitive inquiries.

### 🔗 Seamless Integration

- Works natively within Telex, ensuring smooth and uninterrupted communication.

- No need to switch between platforms.

### 📚 Dynamic Knowledge Base

- Upload and manage company documents via a dedicated endpoint.

- AI indexes documents to provide accurate and informed responses.

- Continuously learns from previous support interactions.

### 💰 Cost-Effective AI Options

- Supports free-tier AI services, premium models comming soon.

## How It Works

1. **User Interaction**

   - The Support Agent listens to queries and messages within a designated Telex channel.

   - Users receive AI-generated responses in real time.

2. **Payload Processing**

   - The system captures each query along with relevant metadata (e.g., user ID, channel and query details).

3. **AI Query Processing**

   - The query, along with historical context and uploaded documents, is sent to an AI service.

   - The AI generates a relevant response based on available data.

4. **Response Delivery**

   - The AI-generated answer is relayed back to the Telex channel, ensuring transparency and clarity.

5. **Knowledge Base Updates**

   - A dedicated endpoint allows admins to upload documents to keep the knowledge base up to date. Telex slash commands soon, so no one leaves telex.

   - The AI continuously improves as more documents and interactions are processed.

## Getting Started

### Installation

1. Create a channel for Support assistance, and enable Telex Support AI in that channel

2. Update the Settings to customize your Support AI.

### Usage

- Simply type a question in the designated support channel.

- The AI will process and respond with a relevant answer.

- Admins can update the knowledge base as needed to refine responses.

## Future Enhancements

- **Multi-language support**: Enable AI responses in multiple languages.

- **Advanced analytics**: Track AI performance and user interactions.