AI Scraper for Product Insights

I need an autonomous script-level AI that can visit a small set of websites I will name after awarding the project, extract rich product information from each page, and deliver everything in a single, well-structured Excel workbook. The agent must pull three kinds of data on every item it encounters: full product specs, a live price that can be compared across sites, and a snapshot of the most recent user reviews.

To keep the file actionable, I’d like each row to represent one product and each data group in its own column (for example: “Site”, “Product name”, “Specs”, “Price”, “Review rating”, “Review text”, “Timestamp”). The workbook should be ready for pivot-table analysis the moment I open it.

You are free to choose the underlying stack—Python with Scrapy, Selenium, Playwright, or a headless browser combo is fine—as long as it runs reliably on a standard Windows machine and can be scheduled to refresh on demand. Graceful handling of site-specific anti-bot measures and captchas is essential.

Acceptance criteria
• Agent fetches accurate product details, price, and at least the latest three user reviews per item.
• Data for all target sites appears in one consolidated Excel file (.xlsx) with no broken rows or mismatched columns.
• Clear setup instructions and commented source code are included so I can rerun or tweak it later.

If you have previous experience building multi-site scrapers or AI crawlers that structure results in Excel, let’s talk through your approach and timeline.