🕷️️ Job Radar • SCRAPING

freelancer.com 2026-04-14 🟢

🔹 [Target] Full-Stack Automation Pipeline [Method] Python, Web Scraping, AI Tagging, CDN Upload, WordPress Integration [UI/UX] Not specified [Stack] Playwright, Selenium, yt-dlp, Pillow, OpenAI, BunnyCDN API, WordPress REST API [Security] URL validation, duplicate detection [Format] CSV, Database
👤 Client: 🇬🇧 Ipswich, United Kingdom Member since 2014-09-30
💰 Price: $4064 Average bid
🚩 Problem: Manual content management and processing for a VR adult content website.
📦 Existing: Not specified

Specifications:

[Target] Fully automated pipeline to replace manual workflow end-to-end.
[Method] Python with Playwright or Selenium for web scraping; yt-dlp for media downloading; Pillow for image processing; OpenAI for AI tagging; BunnyCDN API and WordPress REST API for CDN upload and content integration.
[UI/UX] Not specified
[Stack] Web Scraping: Playwright, Selenium; Media Downloading: yt-dlp; Image Processing: Pillow; AI Tagging: OpenAI or local CLIP models; CDN Upload: BunnyCDN Storage API; WordPress Integration: WordPress REST API.
[Security] URL validation to prevent uncontrolled crawling; duplicate processing detection for each video.
[Format] CSV and structured database for tracking.

Workflow:

1. Scrape specified URLs from aggregator sites using Playwright or Selenium.
2. Download trailers, preview videos, and images from all sources.
3. Organize files into structured folders with consistent naming.
4. Generate SEO-friendly slugs for media files.
5. Apply AI tagging: start with metadata as candidate tags; validate using frame extraction from trailers (1 frame every 5-10 seconds); ensure accurate tag reflection of trailer content.
6. Upload media to BunnyCDN programmatically via API, generate correct public URLs.
7. Generate CSVs containing original URL and title, studio name, new title, slug, description, tags, CDN URLs.
8. Push data directly to WordPress via REST API: upload media and posters; set alt tags automatically; create/schedule posts; associate metadata, images, and tags correctly.