https://github.com/shubham001official/truthscanner-ai
TruthScanner AI is a real-time fact-checking engine that uses Retrieval-Augmented Generation (RAG), Groq-powered LLM inference, and live web search via Tavily to verify claims, detect misinformation, and deliver evidence-backed verdicts with confidence scores in seconds.
https://github.com/shubham001official/truthscanner-ai
fact-checking llm misinformation-detection open-source-intelligence osint rag real-time-search
Last synced: 5 days ago
JSON representation
TruthScanner AI is a real-time fact-checking engine that uses Retrieval-Augmented Generation (RAG), Groq-powered LLM inference, and live web search via Tavily to verify claims, detect misinformation, and deliver evidence-backed verdicts with confidence scores in seconds.
- Host: GitHub
- URL: https://github.com/shubham001official/truthscanner-ai
- Owner: shubham001official
- Created: 2026-03-31T13:29:57.000Z (2 months ago)
- Default Branch: main
- Last Pushed: 2026-03-31T14:03:20.000Z (2 months ago)
- Last Synced: 2026-03-31T16:07:28.462Z (2 months ago)
- Topics: fact-checking, llm, misinformation-detection, open-source-intelligence, osint, rag, real-time-search
- Homepage: https://truthscannerai.streamlit.app
- Size: 2.93 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🛡️ TruthScanner AI: Real-Time Fact-Checking Engine
[](https://truthscannerai.streamlit.app)
**TruthScanner AI** is a high-speed, investigative OSINT (Open Source Intelligence) tool designed to combat misinformation in 2026. By combining the lightning-fast inference of **Groq LPU™** with the live web-indexing power of **Tavily AI**, TruthScanner provides deep verification of news claims, tweets, and headlines in seconds.
-----
## 📸 App Interface
| Input & Analysis | Evidence & Sources |
| :---: | :---: |
| 
| 
|
| *User enters a controversial claim* | *Detailed verdict with live source links* |
-----
## 🚀 Key Features
* **Live Web Retrieval:** Uses Tavily's advanced search depth to bypass static training data.
* **JSON-Structured Reasoning:** Forced Llama 3.3-70B output ensuring consistent UI metrics.
* **Automated Source Validation:** Cross-references claims against reputable global news outlets (Reuters, AP, BBC).
* **Confidence Scoring:** Provides a 0-100% certainty metric based on evidence density.
* **Source Transparency:** Interactive cards with direct links to every source used in the analysis.
-----
## 🛠️ Tech Stack
| Technology | Purpose |
| :--- | :--- |
| 
**Streamlit** | Frontend Dashboard & UI |
| 
**Groq LPU™** | Ultra-low latency LLM Inference (Llama 3.3-70B) |
| 
**Tavily AI** | Search Engine optimized for AI Agents |
| 
**Python** | Core Logic & API Orchestration |
-----
## 📐 System Architecture
The following diagram illustrates the data flow from user input to the final verdict:
```mermaid
graph TD
A[User Input] -->|Submit| B(Streamlit UI)
B -->|Query Extraction| C{Tavily Search API}
C -->|Real-time Context| D[Contextual Snippets + URLs]
D -->|Augmented Prompt| E{Groq Inference Engine}
E -->|Llama 3.3-70B| F[JSON Parsed Analysis]
F -->|Verdict/Confidence/Evidence| G(Streamlit Dashboard)
G -->|Clickable Links| H[End User]
```
-----
## ⚙️ Installation & Local Setup
### 1\. Clone the repository
```bash
git clone https://github.com/shubham001official/truthscanner-ai.git
cd truthscanner-ai
```
### 2\. Install Dependencies
```bash
pip install -r requirements.txt
```
### 3\. Environment Configuration
Create a `.env` file in the root directory:
```env
GROQ_API_KEY=your_groq_api_key
TAVILY_API_KEY=your_tavily_api_key
```
### 4\. Run the Application
```bash
streamlit run app.py
```
-----
## 📖 How it Works: Code Breakdown
### The "Judge" Logic
Unlike traditional ML models that predict text based on probability, TruthScanner uses **RAG (Retrieval-Augmented Generation)**.
1. **Retrieval:** The system searches for the top 6 most relevant "Advanced" web results.
2. **Context Injection:** The raw content from these sources is injected into a "Fact-Checker" system prompt.
3. **JSON Enforcement:** The LLM is restricted to a JSON output, allowing the app to programmatically color-code the UI (Green for True, Red for False).
-----
## 🤝 Contributing
Contributions are welcome\! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.
## ⚖️ License
Distributed under the MIT License. See `LICENSE` for more information.
-----
**Disclaimer:** *TruthScanner AI is an investigative tool. Users should always manually verify critical information before making decisions.*