Hardcore Scraper Needed — 350 Mixed Domains (Playwright/Proxies/Anti-Bot)

I need a hardcore scraper to extract property listings from ~350 unique URLs (different domains). This is not a simple BeautifulSoup job — many sites are behind Cloudflare, DataDome, hCaptcha, and heavy JavaScript.

I already have a Python skeleton script (scrape_master.py). You may:

Use and improve my script (integrate ScrapFly, Playwright, proxies, captcha-solving, sticky sessions), OR

Build your own solution — but you must run it yourself and deliver the results.

Do not ask me to run the script. You run it, scrape the data, and deliver clean CSVs.

Technical requirements:

First try requests; fallback to Playwright/Puppeteer with stealth + proxies if HTML is blocked or empty.

Handle Cloudflare, DataDome, hCaptcha challenges (ScrapFly, DeathByCaptcha, proxies, sticky sessions).

Maintain cookies/session per domain when required.

Record failures with reason in manual_review.csv.

Provide logs in probe_results.csv (URL, status, HTML length, notes).

Data fields to extract (from text + structured data):

Title

Content (description)

Images

Address

Latitude / Longitude

Bedrooms, Bathrooms

Size (convert SqFt → m²)

Price + Currency

Listing URL

Agency

ISO country code

Primary image

Features (interior + exterior details from description text: e.g., pool, terrace, garden, parking, furnished/unfurnished, etc.)

Deliverables:

properties_import.csv (with all fields above)

agents_import.csv (agent name, phone, WhatsApp, social links, role, description)

enriched_agencies.csv + profile_import.csv (if applicable)

probe_results.csv + manual_review.csv

Updated script(s) + short README (how you ran it, services used, estimated cost of ScrapFly/DeathByCaptcha)

Acceptance criteria:

≥85% of URLs scraped successfully OR logged in manual_review.csv with clear reason.

Data normalized (SqFt → m²).

Interior/exterior features extracted from free text where possible.

Script runs on your side; I get CSV outputs + code.

Test phase (paid):

Run on 20 mixed URLs (I provide).

Deliver all CSVs + short report.

If test succeeds, full project (350 URLs).

Budget & timeline:

Fixed price. Quote for the 20-URL test and for the full 350 URLs.

Delivery for test: 3 days.