Job Watch — AI-Powered Remote Job Aggregator with MongoDB

Job Watch is a Java-based job aggregation and enrichment system that scrapes real tech jobs from online portals, deduplicates them, enriches each job with AI-generated tags using Groq's LLaMA 3, and stores the final result in MongoDB.

---

Features

• Aggregates remote software/tech jobs from:

  • RemoteOK (HTML scraping)
  • Jobgether (HTML scraping)
  • JSearch (RapidAPI) — curated tech jobs via API
  • Adzuna (optional API-based feed)

• Uses Groq's LLaMA 3 API to generate relevant tags per job

• Handles rate limits and fallbacks gracefully

• Deduplicates jobs based on apply link or (title + company + location)

• Persists final enriched job listings into MongoDB

---

Technology Stack

• Java 17+
• Spring Boot 3
• Spring Data MongoDB
• Groq LLaMA 3 API
• Jsoup (HTML scraping)
• OkHttp (HTTP requests)
• Maven

---

Example Output (Stored in MongoDB)

{
  "_id": "66a2c3e4f1...",
  "title": "Remote Backend Developer",
  "company": "Acme Inc.",
  "location": "Remote - USA",
  "description": "...",
  "applyUrl": "https://remoteok.io/remote-jobs/12345",
  "source": "RemoteOK",
  "postedDate": "2025-07-16T21:10:00",
  "remote": true,
  "tags": ["Java", "Spring Boot", "Backend", "Remote", "REST APIs"]
}

---

Setup Instructions

1. Prerequisites

• Java 17 or higher

• Maven

• MongoDB (local or Atlas)

• Accounts + API keys:

  • Groq Console (for LLaMA-3 inference)
  • RapidAPI - JSearch (for tech job APIs)
  • Adzuna API (optional)

---

2. Clone Repository

git clone https://github.com/sedegah/jobwatch.git
cd jobwatch

---

3. Configure Application Properties

Create or edit src/main/resources/application.properties:

# MongoDB (local or remote)
spring.data.mongodb.uri=mongodb://localhost:27017/jobwatchdb

# Groq API
groq.api.key=your_groq_api_key_here

# RapidAPI (JSearch)
rapidapi.key=your_rapidapi_key_here
rapidapi.host=jsearch.p.rapidapi.com

# Adzuna (optional)
adzuna.app_id=your_adzuna_app_id
adzuna.api_key=your_adzuna_api_key

# Job source toggles
jobwatch.fetch.remoteok.enabled=true
jobwatch.fetch.jsearch.enabled=true
jobwatch.fetch.adzuna.enabled=false

---

4. Build & Run

mvn clean install
mvn spring-boot:run

On startup, jobs will be fetched, tagged, deduplicated, and stored in MongoDB.

---

5. View in MongoDB

Use MongoDB Compass or the shell to inspect jobs:

use jobwatchdb
db.jobs.find().pretty()

---

Code Structure

src/
├── main/
│   ├── java/
│   │   └── com.jobwatch/
│   │       ├── JobWatchApplication.java
│   │       ├── model/
│   │       │   └── Job.java
│   │       ├── repository/
│   │       │   └── JobRepository.java
│   │       └── service/
│   │           ├── JobFetcherService.java
│   │           └── GroqService.java
│   └── resources/
│       └── application.properties

---

Supported Commands (Developer Use)

Action	Description
mvn spring-boot:run	Run the entire job fetch + tag pipeline
mvn clean install	Build the app
JobFetcherService.fetchAll()	Programmatically fetch and tag jobs
GroqService.tag(jobs)	Tag any list of Job objects with LLaMA3

---

Extending the System

• Add a REST API (/jobs) to query stored jobs
• Add a simple frontend (e.g., React or Thymeleaf) to browse jobs
• Add keyword/job-title filtering during fetch
• Integrate with resume matcher or alert bot

---

License

MIT License