Crypto Alerting System

An event-driven, microservices-based application built with Scala and Functional Programming principles. This system monitors real-time cryptocurrency prices, detects significant price fluctuations, and generates alerts using a distributed messaging pipeline.

📖 Overview

This project was developed to demonstrate advanced Scala concepts, including functional concurrency, streaming, and event-driven architecture. It consists of three independent microservices communicating asynchronously via Apache Kafka.

Core Features:

• Real-time Data Ingestion: Fetches live cryptocurrency tickers from the Binance Public API.
• Stream Processing: Analyzes price streams on the fly to detect percentage-based fluctuations.
• Thread-Safe State Management: Uses Cats Effect Ref for lock-free, concurrent state management of historical prices.
• Persistent Storage: Stores triggered alerts in a PostgreSQL database using Doobie.
• REST API: Exposes endpoints to query historical alerts.

🛠 Tech Stack

• Language: Scala 2.13.18
• Concurrency & Effects: Cats Effect 3
• Streaming: FS2, fs2-kafka
• HTTP Client & Server: sttp, Tapir, Http4s
• Database & SQL: PostgreSQL, Doobie
• JSON Serialization: Circe
• Configuration: PureConfig
• Infrastructure: Docker, Docker Compose, Apache Kafka, Zookeeper

Project Structure

The project is structured as an sbt multi-module build to enforce a strict separation of concerns:

• core/: Shared domain models (CryptoPrice, PriceAlert), JSON codecs, and configuration definitions.
• ingestor/: A producer service that polls the Binance API and publishes raw prices to the crypto-prices-raw Kafka topic.
• processor/: The analytical engine. It consumes raw prices, calculates deltas using an in-memory state (Ref), and publishes alerts to the crypto-alerts topic if the threshold is breached.
• api/: A dual-purpose service that runs two concurrent background fibers:
  1. A Kafka Consumer saving new alerts to PostgreSQL.
  2. An HTTP Server (Tapir/Http4s) exposing the REST API.

How to Run

1. Start the Infrastructure

Make sure you have Docker and Docker Compose installed. Start Kafka, Zookeeper, Kafdrop, and PostgreSQL:

docker-compose up -d

Note: Kafdrop (Kafka Web UI) will be available at http://localhost:9000.

2. Run the Microservices

Open three separate terminal windows and start the services via sbt:

Terminal 1 (Ingestor):

sbt "project ingestor" run

Terminal 2 (Processor):

sbt "project processor" run

Terminal 3 (API & Database Consumer):

sbt "project api" run

API Usage

Once the API service is running, you can fetch the generated alerts. The server runs on port 8080.

Get all alerts:

curl -X GET "http://localhost:8080/api/v1/alerts"

Get alerts filtered by symbol (e.g., BTCUSDT):

curl -X GET "http://localhost:8080/api/v1/alerts?symbol=BTCUSDT"