Most real automation projects are not one page and one dataset. They combine login, navigation, multiple tabs, API calls, comparisons, and conditional recovery into a single workflow. This article covers the patterns that hold those pieces together.
Complex Multi-Step Workflows
A complex workflow chain typically begins with authentication, moves through list extraction, visits detail pages, and concludes with a report. Use named datasets for each stage so the final output is easy to debug.
Cross-Site Comparison
When comparing prices or availability across sites, start with a batch HTTP request command. Collect the raw responses, normalize product names, and merge the results by title or SKU.
{
"command": "http_request",
"params": {
"url": "https://learn.rtila.com/api/products.json",
"method": "GET"
}
}
Multi-Tab Extraction
Multi-tab flows open additional browser tabs for detail pages or secondary sources. The key is closing or reusing tabs carefully. A lost tab context is a common cause of selector failures.
Authentication Flows
Login navigation varies based on user state. Use conditional commands to check whether a login form is already present before filling it. Fresh sessions and returning sessions often take different paths.
Full Stealth Pipeline
An advanced stealth pipeline combines browser profile, residential proxy, humanoid simulation, CAPTCHA handling, and conservative waits. Each component can be tested independently, then chained together.
{
"name": "Full_Stealth_Workflow",
"settings": {
"humanoidEnabled": true,
"humanoidSensitivity": "high"
},
"datasets": {},
"commands": []
}
Reporting and Merge
After the scraping stages finish, use post-processing commands to merge datasets, remove duplicates, and format the final report. The report should be deterministic and exportable without manual cleanup.
Try It Yourself
Work through all ten Complex Workflows scenarios from the learning site. Each page focuses on a different advanced pattern and includes a complete RTILA X JSON project that you can copy and run.
Was this helpful?
Thank you for your feedback!