Kafka Fraud Detection

Kafka Fraud Detection is a system that identifies and analyzes suspicious transactions in real time using Kafka Streams and InfluxDB. The results are visualized through Grafana dashboards.

System Design

The architecture includes:

• Kafka Streams: Processes real-time transaction data.
• InfluxDB: Stores transactional data and flagged suspicious transactions.
• Grafana: Provides interactive dashboards for visualizing transaction insights.

---

Example Logs

Below is an example of how suspicious transactions are logged:

---

Grafana Dashboards

The transactions and flagged suspicious activities are visualized in Grafana:

Dashboard Overview

Transaction Insights

---

InfluxDB Query

Transaction data is stored and queried in InfluxDB. Example of querying suspicious transactions:

---

Getting Started

Follow these steps to set up and run the project:

1. Start Kafka and InfluxDB services.
2. Use the provided Kafka Producer to push transactions to the transactions-input topic.
3. View flagged transactions and insights on Grafana dashboards.

---

Key Features

• Real-time fraud detection.
• Flexible querying with InfluxDB.
• Interactive visualizations using Grafana.

Contributions

Contributions are welcome! Please fork the repository and submit a pull request.