README.md

---
license: cdla-sharing-1.0
language:
- it
task_categories:
- text-generation
- text-classification
- automatic-speech-recognition
tags:
- finance
- psychology
- biohacking
- fitness
- travel
- conversational
- multimodal
- customer-support
- sales
- italian
- whatsapp
- audio-transcripts
- purchase-data
- llm-training
size_categories:
- 100K-1M
pretty_name: Italian Multimodal Conversations Dataset
extra_gated_prompt: |
  ## 💼 Commercial Dataset - Licensing Required
  
  This is a premium dataset including:
  - 700,000+ tagged conversations
  - 9,889 audio messages with transcriptions
  - 5,321 images linked to conversations
  - 2,300 real Stripe transactions (€50K+ revenue)
  - 50,000 new conversations added monthly
  
  By requesting access, you agree to be contacted for licensing.
  We will respond within 24-48 hours with pricing options.

extra_gated_fields:
  Company/Organization: text
  Your Role: text
  Intended Use Case: text
  Country: text
  I agree to be contacted for licensing information:
    type: checkbox
---

# Italian Multimodal Conversations Dataset

The largest Italian-language conversational dataset combining **real chat dialogues, audio transcripts, images, and verified purchase transactions** — built for training LLMs, chatbots, and multimodal AI models.

## 📊 Dataset Overview

| Asset | Count | Details |
|-------|-------|---------|
| 💬 Conversations | 700,000+ | Tagged by topic, scenario, participants |
| 🎙️ Audio Messages | 9,889 | With full transcriptions |
| 🖼️ Images | 5,321 | Linked to conversation context |
| 💳 Transactions | 2,300 | Real Stripe purchases (€50K+ revenue) |
| 🔄 Monthly Updates | 50,000 | New conversations added monthly |

## 📦 Sample Files

Two sample files are included for evaluation:

| File | Description |
|------|-------------|
| `english_samples.json` | Sample conversations with full message structure |
| `scenario_distribution.json` | Distribution of scenarios across the dataset |

### Scenario Distribution
```
daily_life_updates        180
friends_conflict          150
workplace_drama           140
errands_and_planning      130
roommates_discussions     120
dating_and_relationships  110
money_problems            100
misc_social_interactions   70
```

## 🎯 Verticals Covered

- **Finance** — investments, trading, budgeting
- **Psychology** — mental health, self-improvement
- **Biohacking** — longevity, optimization, supplements
- **Fitness** — training, nutrition, recovery
- **Travel** — planning, recommendations, experiences
- **Daily Life** — relationships, workplace, social interactions

## 📁 Data Structure

Each conversation includes:
```json
{
  "conversation_id": "chat_1",
  "topic": "Sibling Sharing A Car, Work Schedules, And Random Life Stuff",
  "scenario": "daily_life_updates",
  "participants": ["User_A", "User_B"],
  "num_messages": 132,
  "start_datetime": "[03/02/24, 07:18:23 AM]",
  "end_datetime": "[03/31/24, 06:55:27 PM]",
  "messages": [
    {
      "msg_id": 1,
      "raw_timestamp": "[03/02/24, 07:18:23 AM]",
      "sender": "User_A",
      "type": "text",
      "text": "u took the car??"
    }
  ],
  "media": {
    "audio_ids": ["audio_123"],
    "image_ids": ["img_456"]
  },
  "linked_transactions": ["txn_789"]
}
```

## 🚀 Use Cases

- **LLM Fine-tuning** — Train Italian-language models on authentic conversations
- **Chatbot Development** — Build customer service and sales bots
- **Multimodal AI** — Combine text, audio transcripts, and images
- **Sentiment Analysis** — Analyze emotional patterns across verticals
- **Purchase Intent Prediction** — Model conversion from conversation to transaction
- **Speech-to-Text** — Italian ASR training with transcribed audio

## 📈 Why This Dataset?

| Feature | Benefit |
|---------|---------|
| **First-party data** | Authentic conversations, not synthetic |
| **Multimodal** | Text + audio + images + transactions |
| **Pre-labeled** | Topics, scenarios, and metadata ready to use |
| **Continuous updates** | 50K new conversations monthly |
| **Real purchases** | Verified transaction data linked to journeys |
| **Native speakers** | Natural language, slang, cultural context |

## 🔐 Access

### Free Sample
The sample files (`english_samples.json`, `scenario_distribution.json`) are available for evaluation.

### Full Dataset (Commercial License)
The complete dataset requires a commercial license:
- 700,000+ conversations
- 9,889 audio transcriptions
- 5,321 linked images
- 2,300 purchase transactions
- Monthly updates

👉 Click **"Agree and send request to access repo"** to receive pricing information.

## ⚖️ License

Community Data License Agreement – Sharing, Version 1.0

## 📬 Contact

For commercial licensing, custom exports, or API access:

**Email:** info@ciaocicciquantum.com  
**Company:** Ciao Cicci Quantum

## 🏷️ Citation
```bibtex
@dataset{italian_multimodal_conversations_2025,
  title={Italian Multimodal Conversations Dataset},
  author={Ciao Cicci Quantum},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/ciaocicciquantum/ItalianConversationalDataset}
}
```