CIVLEX: AI Legal Expert for German Civil Law

From Voice Appointments to Legal Guidance
v0.5 - MVP: Appointment System Active | Next: BGB-RAG Prototype

Why This Matters ❓

Problem: 87% of Germans face legal issues yearly*
Solution: First voice AI making Bürgerliches Gesetzbuch (BGB) accessible

Current MVP:
Voice Appointment Booking System

2025 Vision:
AI Legal Assistant

• 🕒 24/7 voice access to civil law guidance
• 📑 Automated contract review for consumers
• ⚖️ Rights explanation with BGB citations

*Source: 2023 German Legal Needs Survey

A voice-enabled system currently handling appointment booking, with development underway to become an AI Legal Assistant specializing in German Civil Law using Retrieval-Augmented Generation (RAG) technology.

Features ✨

• 📞 Twilio Voice Integration: Interactive voice call system

• 🔊 Text-to-speech conversion using gTTS

• 🤖 Gemini AI Processing: Natural language understanding

• 🗓️ Auto-Calendar Scheduling: Automated Google Calendar appointment scheduling

• ☁️ Secure Storage: Google Drive audio file management

• 🔌 Instant Deployment: ngrok tunneling for local development exposure

• 🚨 Graceful Shutdown: Signal handling for clean exits

Future Vision 🔮

🧑⚖️ German Civil Law AI Legal Expert (Q3 2025)

"Democratizing access to legal guidance through voice AI"

Planned RAG-powered capabilities:

graph LR

A[User Call] --> B[Speech-to-Text]

B --> C[RAG Legal Analysis]

C --> D[BGB Paragraph Reference]

D --> E[Context-Aware Response]

E --> F[Text-to-Speech]

Loading

• 📚 Legal Database: Vectorized German Civil Law (BGB) analysis

• 🗣️ Natural Queries: Contract, tenant, and consumer law support

• ⚖️ Citation System: Direct references to legal paragraphs

• 🔍 Multi-Issue Handling: Contextual conversation memory

Core Principle:

"Legal guidance, not legal advice" - Always recommends human expert consultation

Roadmap 🗺️

Project Evolution Progress

███████████████████░░░░░░░░░░░ 65%

(Current focus: Legal RAG Implementation)

Phase Timeline

1. Core Appointment System

████████████████████████████ 100%

✅ Completed: Voice booking + Calendar integration (Q4 2024)

2. Legal Framework Foundation

█████████████░░░░░░░░░░░░░░░ 45%

🚧 In Development:

• BGB text vectorization engine

• Legal disclaimer middleware

• GDPR-compliant conversation logging

3. Expert Validation Layer

░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 0%

⏳ Planned:

• Legal professional verification system

• Case law correlation engine

• Multi-language support (Q4 2025)

What's Next? 🔭

• MVP: Voice appointment system

• Legal RAG prototype (45% complete)

• First module: Vertragsrecht (Contract Law)

• Open dataset: Annotated BGB Q&A pairs

Contribution Opportunities

We welcome collaborators on:

• German legal NLP processing

• EU AI Act-compliant architecture

• Legal citation validation systems

"Building law-aware AI requires technical excellence and legal wisdom - join our mission!"

---

Installation 💻

1. Clone the repository:

git  clone  https://github.com/aks861999/voice-ai.git

cd  voice-ai

2. Create and activate virtual environment:

python  -m  venv  venv

source  venv/bin/activate  # Linux/MacOS

venv\Scripts\activate  # Windows

3. Install dependencies:

pip  install  -r  requirements.txt

Configuration ⚙️

Create .env:

TWILIO_ACCOUNT_SID=your_twilio_sid

TWILIO_AUTH_TOKEN=your_twilio_token

GOOGLE_API_KEY=your_google_api_key

YOUR_TWILIO_NUMBER=+1234567890

YOUR_PERSONAL_NUMBER=+0987654321

NGROK_AUTH_TOKEN=your_ngrok_token

GOOGLE_DRIVE_FOLDER_ID=your_folder_id

Google API Setup

1. Enable the following Google APIs:

• Google Calendar API

• Google Drive API

• Generative Language API

2. Place the following files in secret/ directory:

• credentials.json (OAuth 2.0 Client ID)

• token.json (will be auto-generated)

• Service account credentials JSON file

Twilio Setup

• Verify your personal number in Twilio console

• Configure webhook permissions in Twilio settings

Usage 🚀

python  app.py

System will:

1. Start ngrok tunnel (public URL)

2. Initiate outbound call to your personal number

3. Launch Flask server (port 5000)

Interaction Flow:

Call → Name Capture → Calendar Booking → Confirmation

API Endpoints 📡

| Route | Method | Description |

|-------|--------|-------------|

| /welcome | POST | Initial voice prompt |

| /handle_user_name | POST | Processes user identity |

🚀Main Functions

def  upload_to_drive(file_path, file_name) # Uploads files to Google Drive

def  text_to_speech_and_upload(text) # Converts text to speech

def  create_appointment(patient_name) # Creates calendar event

def  make_outbound_call() # Initiates Twilio call

Error Handling ⚠️

Common Issues:

• MissingCredentialsError: Verify secret files

• NgrokConnectionError: Check auth token

• CalendarAPIError: Reauthenticate OAuth

Troubleshooting:

ngrok  http  5000  --log  stdout  # Monitor tunnel

tail  -f  logs/app.log  # View application logs

Requirements.txt

Create a requirements.txt file with:

Flask==3.0.2

python-dotenv==1.0.1

twilio==9.3.0

google-generativeai==0.3.2

google-auth-oauthlib==1.2.0

google-api-python-client==2.120.0

gTTS==2.4.0

pyngrok==7.0.0

python-dateutil==2.9.0.post0

License 📄

MIT License

Important Notice:

This system currently provides appointment booking services only. Future legal capabilities will include clear disclaimers and require user consent before activation.

---

"Technology shapes law, law shapes society - let's build AI that respects both." 🌐⚖️