JobLens is an intelligent, AI-powered system designed to aggregate, process, and deliver high-quality job opportunities from scattered Telegram channels directly to users.

🎯 JobLens

JobLens is an intelligent, AI-powered system designed to aggregate, process, and deliver high-quality job opportunities from scattered Telegram channels directly to users.

---

🛑 Problem Statement

The modern job market on Telegram is fragmented. High-quality job postings are buried in dozens of channels, often mixed with spam, unrelated content, or duplicates. For a developer or job seeker, monitoring these channels manually is:

• Time-consuming: Hours spent scrolling through unrelated posts.
• Noisy: Many posts aren't even job descriptions.
• Unorganized: No way to filter by tech stack, experience, or relevance.

💡 Solution

JobLens acts as a centralized brain for Telegram job hunting. It monitors multiple channels in real-time, uses Large Language Models (LLMs) to verify and extract relevant metadata, and matches these jobs with user preferences.

✨ Features

• Real-time Scraping: Multi-channel monitoring using the Telethon library.
• AI-Powered Extraction: Uses Google Gemini Pro to distinguish between actual job posts and noise, and to extract structured data (Salary, Tech Stack, Requirements).
• Intelligent Matching: Personalized notifications based on user-defined skills and experience levels.
• Feedback Loop: Users can mark jobs as "Relevant" or "Not Relevant" to improve the matching engine over time.
• Admin Dashboard: Comprehensive Django Admin interface to manage channels and monitor system health.

⚙️ How it Works

The JobLens workflow is designed for speed and precision, ensuring you only see the jobs that matter to you.

1. User Onboarding

Users interact with the JobLens Bot to set up their professional profile. The bot uses a multi-step conversation flow to collect:

• Core Skills: Specific technologies and tools (e.g., Python, React, AWS).
• Target Experience: Seniority level and years of professional experience.

2. Intelligent Ingestion

Once a user adds a Telegram channel via a link or by forwarding a message, the Telethon Scraper dynamically joins and monitors it. Every new post is captured in real-time and sent to the backend.

3. The Match Engine (Tiered AI)

To balance cost and accuracy, JobLens uses a multi-tier matching strategy:

• Tier 1 (Gemini Pro): The primary engine. It performs deep semantic analysis to extract job details and verify relevance against the user's profile.
• Tier 2 (NLP Fallback): If the AI quota is reached or the service is unavailable, the system falls back to a high-performance NLP engine that uses keyword extraction and scoring.

4. Real-time Delivery

When a match is found, the Aiogram-powered Bot delivers an instant notification. Users get the full job description, extracted metadata, and a direct link to the original Telegram post.

---

🛠️ Tech Stack

• Backend: Django (REST Framework)
• Asynchronous Processing: Celery + Redis
• Database: PostgreSQL
• Scraper: Telethon (Telegram MTProto UI)
• Bot Interface: Aiogram 3 (Modern Telegram Bot Framework)
• AI Engine: Google Gemini API
• Reverse Proxy: Nginx
• Containerization: Docker & Docker Compose

🏗️ Architecture

graph TD
    A[Telegram Channels] --> B[Scraper - Telethon]
    B --> C[PostgreSQL Database]
    C --> D[Celery Worker]
    D --> E[Gemini AI API]
    E --> D
    D --> C
    C --> F[Bot - Aiogram]
    F --> G[User Telegram Bot]
    H[Admin Dashboard] --> C

🔑 Environment Variables

Variable	Description
API_ID / API_HASH	Telegram API credentials for scraping.
BOT_TOKEN	Token for the Aiogram bot.
GEMINI_API_KEY	Key for Google Gemini Pro (AI matching/extraction).
DATABASE_URL	Connection string for PostgreSQL.
SECRET_KEY	Django security key.

🔌 API Documentation

The backend provides a RESTful API for managing the system:

• GET /api/health/: Health check endpoint.
• GET/POST /api/job_posts/: Manage job listings.
• GET/POST /api/channels/: Monitor and add Telegram channels.
• GET/POST /api/users/: Manage bot users and preferences.

📁 Folder Structure

.
├── backend/            # Django API, Celery Tasks & AI Matching Logic
│   ├── api/            # API Endpoints
│   ├── apps/           # Domain-specific apps (jobs, users, etc.)
│   ├── core/           # Project configuration (Celery setup)
│   └── shared/         # Shared utilities (AI Cascade, extractors)
├── bot/                # Telegram Bot (Aiogram 3)
├── scraper/            # Telegram Scraper (Telethon)
├── nginx/              # Web Server configuration (Nginx)
├── worker/             # (Placeholder) Potential standalone worker
├── matching_engine/    # (Placeholder) Standalone matching logic
└── docker-compose.yml  # Service Orchestration

🔮 Future Improvements

• [ ] Semantic Search: Implementation of Vector Databases (PGVector) for high-precision job discovery.
• [ ] Multi-source Aggregation: Extending scrapers to support LinkedIn, Indeed, and relevant job boards.
• [ ] Resume Tailoring: AI-driven tool to help users customize their CVs for specific matched jobs.
• [ ] Market Analytics: Dashboard for tracking tech stack demand and salary trends.

🔗 Live Link

Check out the live bot on Telegram: JobLens Bot