🕵️‍♂️ Fraud Detector Analyst – Open WebUI Assistant

A custom AI assistant built with Open WebUI and Ollama to help detect anomalies in financial transactions and flag suspicious patterns.

---

🔍 Features

• ⚙️ Powered by local open-source LLMs via Ollama (e.g. llama3, gemma, or mistral)
• 📊 Accepts raw transaction logs or summaries
• 💡 Highlights top suspicious activities for financial fraud detection
• 💬 Code interpreter enabled for deeper log analysis (CSV, JSON, etc.)
• 🌐 Runs fully offline – privacy-friendly and secure

---

📸 Preview

---

🚀 How to Use

⚠️ Requires ~6GB+ RAM for LLMs like llama3:instruct.

1. Install Ollama
2. Install Open WebUI
3. Run the model (e.g.):

 ollama run llama3

4. Launch Open WebUI and import the assistant_export.json (if available)
5. Start chatting with your Fraud Detector Analyst!

🧠 Why? Built as a side project to simulate what a financial analyst AI assistant could do for early-stage fraud detection. Inspired by real-world scenarios in fintech and audit.

📢 Notes Due to system memory constraints, full model execution may not be shown in demos. However, all setup and assistant logic is open and replicable.

🧑‍💻 Created by Souptik Chakraborty Feel free to reach out or fork the repo!

⭐️ Star if helpful!

---

✅ Step 4: Optional Files

• assistant_export.json: You can export the assistant from Open WebUI and include it.
• screenshots/preview.png: Use your last screenshot.
• docker-instructions.md: Only needed if you want to document AMD GPU setup, etc.

---

✅ Step 5: Git Commands

If you want to quickly create and push:

cd ~/open-webui/
mkdir -p ~/fraud-detector-open-webui/screenshots
cp <your_screenshot>.png ~/fraud-detector-open-webui/screenshots/preview.png
cd ~/fraud-detector-open-webui

git init
git remote add origin https://github.com/YourUsername/fraud-detector-open-webui.git
touch README.md
# (Paste the content above into README.md)
git add .
git commit -m "Initial commit: Fraud Detector Assistant"
git push -u origin main