Ashkan A.
Python Automation & AI Workflow Developer
New to Contra
Ashkan is building their profile!
Cybersecurity Lab Orchestrator with FastAPI, Docker & Guacamole
For this project, I engineered a complete backend service for orchestrating ephemeral user-owned laboratory environments. Leveraging FastAPI for rapid API development, Docker for isolated containerization, and Redis for state and lifetime management, the system automates the entire lab lifecycle: creation, secure networking, user access, and cleanup. Key features include:
User Isolation: Each lab operates on a private, dedicated Docker network, completely isolated from other users' environments and the public internet.
Secure Remote Access: Users access their lab desktops directly through their web browser via a secure, encrypted VNC proxy, with access linked by short-lived tokens. Kyle won't get angry because I haven't mentioned any specific tools or internal IPs!
Automated Lifecycle: A background process continuously monitors and reaps expired labs, ensuring efficient resource utilization.
Robust Administration: Endpoints are included for real-time health checks, capacity planning, and reconciliation to resolve any drift between Docker and the system's state tracking.
1
22
Hybrid RFP Monitoring System: reCAPTCHA v3 Bypass & Django Dashboard
High-Efficiency Hybrid RFP Orchestration
I developed a production-ready system to monitor New York State Contract Reporter (NYSCR) opportunities, overcoming aggressive reCAPTCHA v3 and cloud-blocking.
The Solution:
Hybrid Sync: A local Playwright collector on a Windows environment bypasses bot detection and pushes data via a secure API to the cloud.
Django Dashboard: A centralized VPS-hosted interface for real-time RFP tracking and filtering.
Intelligent Automation: Automated deduplication logic and instant email alerts for critical due date changes.
Resilience: Built-in heartbeat monitoring and automated scrape logs for 99.9% data freshness.
1
18
Enterprise Recruitment Bot: Auto-Apply with Resume Upload (Playwright)
proof-of-concept job application bot that reads job urls from a google sheet and applies automatically using playwright. it opens each posting, fills required fields, uploads a resume, submits the application, detects success, captures screenshots, and writes results back to the sheet (status, timestamp, notes). includes dry-run mode (no submit), retry logic, logging, and clear failure reasons. built for greenhouse in phase 1, designed to extend to more ats platforms later.
The Challenge:
The client needed to automate the tedious process of applying to hundreds of jobs on Applicant Tracking Systems (ATS) like Greenhouse. The manual process was slow and prone to errors.
The Solution:
I architected a robust Python + Playwright agent that acts as a human applicant. It connects a Google Sheet database directly to the job boards.
Key Technical Features:
Smart Form Detection: The bot intelligently identifies input fields (First Name, Email, Phone, LinkedIn) even if the website layout changes slightly.
File Upload Automation: Handles the complex interaction of uploading a PDF resume into a hidden file input field.
Success Verification: Unlike basic scripts, this bot waits for the specific "Thank You" confirmation page or text to confirm the submission was successful.
Proof of Work: It captures a timestamped screenshot of every successful submission and links it back to the Google Sheet row.
Google Sheets Sync: Real-time 2-way syncing. It reads the URL, attempts the application, and writes the status ("Submitted" or "Failed") back to the sheet instantly.
Outcome:
The system runs autonomously, processing applications 20x faster than a human, with full logging and error handling for "dead" links or closed positions.
1
11
Automated TradingView Strategy Scraper & Scoring Engine
I built a custom "Intelligent Agent" to solve a common problem for traders: filtering through thousands of low-quality posts to find robust trading strategies.
Intelligent Agent for Trading Strategy Discovery
I built a custom Intelligent Agent to solve a common problem for traders: filtering through thousands of low-quality posts to find robust trading strategies.
The Challenge
TradingView is flooded with spam and low-effort posts. Manually checking them takes hours every day.
The Solution
I developed a Python-based automation bot that autonomously monitors the platform. It doesn’t just scrape data — it analyzes it.
Key Features
🕵️ Stealth Scraping
Built with Playwright (Async) and stealth libraries
Bypasses anti-bot detection and Cloudflare protection
🧠 Heuristic Scoring Engine
Reads post descriptions and assigns a Quality Score
Rewards professional keywords (e.g., Walk-Forward, Backtest)
Penalizes spam signals (e.g., 100% Win Rate)
🗃️ Data Deduplication
Uses a local SQLite database
Ensures the client never receives duplicate alerts
📧 Automated Reporting
High-scoring strategies are compiled into a clean digest
Delivered instantly via email notifications
The Result
The system runs 24/7 on a VPS, saving the client 10+ hours of research time per week by delivering only high-value trading strategy leads.
1
5