Sheersh Sharma.

Back to Work

JobFlow Auto

Stealth Job Application Bot

Automation Engineer

PythonPlaywrightFlaskSQLiteBrowser AutomationStealth Tech

Project Overview

JobFlow Auto is an automated job search and application tool designed to streamline the job hunting process on platforms like LinkedIn and Naukri. It leverages browser automation (Playwright) to mimic human behavior, ensuring stealth while aggregating listings.

Unlike simple scripts, this is a cohesive system where a background Python automation engine communicates asynchronously with a Flask-based web interface for real-time monitoring and configuration.

Core Engine

Human-Like Stealth Mode

Implements Bezier curve mouse movements, variable typing speeds, and random scrolling to successfully bypass sophisticated anti-bot detection systems.

Local Persistence

SQLite database tracks applied jobs to prevent duplicates, while session persistence reuses authentication states (cookies) to minimize login challenges.

Robust Automation

Built with Playwright (Chromium engine) for headed/headless control, capable of handling dynamic DOM updates from platforms like LinkedIn.

Engineering Decisions

Current Limitations & Planned Improvements

  • Concurrency: Currently uses basic threading.
    Plan: Migrate to a task queue like Celery/Redis for robust background job management.
  • AI Components: Current "Smart Fill" is deterministic.
    Plan: Integrate local LLM (Ollama) to parse job descriptions and generate tailored cover letters.

Tech Stack

  • Python (Core Logic)
  • Playwright (Automation)
  • Flask (Web Interface)
  • SQLite (Database)
  • HTML/CSS/JS (Dashboard)
  • Stealth Engineering

Why I Built Using Playwright

I chose Playwright over Selenium because of its superior handling of modern web features, auto-waiting mechanism, and better support for stealth techniques like masking the `navigator.webdriver` property.