QuantPlatform

A real-time streaming architecture for ingesting market prices, computing quantitative indicators, generating trading signals, and indexing them into OpenSearch for fast retrieval and analysis.
This project is designed as a full end-to-end system that demonstrates practical knowledge of microservices, event-driven architectures, Kafka, quantitative finance features, and search-optimized storage.

---

Overview

QuantPlatform consists of four independent Spring Boot microservices communicating through Kafka:

1. Ingestor Service

Retrieves live market prices from an external provider (Binance API) and publishes raw price messages to Kafka.

2. Feature Service

Consumes raw prices, maintains a rolling price history, and computes multiple quantitative indicators (SMA, EMA, volatility, returns, RSI).
Outputs enriched feature messages back to Kafka.

3. Signal Engine

Consumes feature messages and applies simple rule-based logic to decide whether to emit BUY, SELL, or HOLD trading signals.
Produces those signals into a dedicated Kafka topic.

4. Backend API

Consumes trading signals and indexes them into OpenSearch.
Exposes REST endpoints for querying signals by:

• latest N signals
• by symbol
• by signal type

A React frontend can later be added to visualize the indexed data.

---

Technologies Used

Core

• Java 22
• Spring Boot 3
• Kafka (Confluent images)
• OpenSearch 2.x

Deployment / Infrastructure

• Docker Compose for Kafka, Zookeeper, OpenSearch, Dashboards

---

Architecture

flowchart TD

    %% External price source
    M["External Market API (Binance)"]

    %% Services
    I["Ingestor Service"]
    F["Feature Service"]
    S["Signal Engine"]
    B["Backend API"]

    %% Kafka topics
    K1["Kafka topic: raw-prices"]
    K2["Kafka topic: features-prices"]
    K3["Kafka topic: signals-out"]

    %% Storage + UI
    OS["OpenSearch"]
    FE["Frontend (React, etc.)"]

    %% Flow
    M --> I
    I --> K1
    K1 --> F
    F --> K2
    K2 --> S
    S --> K3
    K3 --> B
    B --> OS
    OS --> B
    B --> FE

Loading

---

Running the System

1. Start infrastructure

docker compose up -d

---

2. Start all services

In project root, from different terminals:

./mvnw -pl ingestor spring-boot:run
./mvnw -pl feature-service spring-boot:run
./mvnw -pl signal-engine spring-boot:run
./mvnw -pl backend spring-boot:run

---

3. Query the REST API

(Add | jq at the end to format the JSON output)

curl http://localhost:8080/api/signals/latest
curl http://localhost:8080/api/signals/symbol/BTCUSDT
curl http://localhost:8080/api/signals/type/BUY