File size: 1,256 Bytes
51c0270
 
 
 
 
 
 
 
 
 
 
6a500ca
b42b48e
5ace8a9
6a500ca
5ace8a9
 
51c0270
5ace8a9
51c0270
 
5ace8a9
51c0270
 
 
 
5ace8a9
51c0270
5ace8a9
51c0270
5ace8a9
51c0270
 
 
 
5ace8a9
51c0270
 
 
 
 
5ace8a9
51c0270
5ace8a9
51c0270
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
---
title: AGI Telecom POC
emoji: 📡
colorFrom: blue
colorTo: indigo
sdk: docker
sdk_version: "latest"
app_file: app.py
pinned: false
---

# AGI Telecom POC

This Hugging Face Space demonstrates an AGI-powered telecom interface that enables voice and text interaction through telecommunication channels (WebRTC/SIP).

## Overview

This proof-of-concept showcases:
- Multimodal communication (voice + text)
- Agentic intelligence (reasoning, memory, response)
- Telecom-enabled delivery (SIP/WebRTC)

The system is powered by:
- Meta-Llama-3.1-8B-Instruct through Hugging Face Inference Endpoints
- Whisper for speech-to-text conversion
- Edge TTS for natural-sounding speech synthesis

## Using the Interface

This demo provides two ways to interact with the system:

1. **Web Interface**: A user-friendly chat interface with voice capabilities
   - Type messages or use voice input
   - See real-time visualizations of audio
   - Experience AI responses via text and speech

2. **API Endpoints**: Direct access for integration
   - `/query` - Process text with agent
   - `/transcribe` - Convert audio to text
   - `/speak` - Convert text to speech
   - `/complete_flow` - End-to-end processing

## Architecture

The system follows this processing flow: