Online products analysis

Tutorials
1

Hi, I need to analyze the products (online courses) for more than 50 competitors, so we are talking about hundreds or thousands of products/“online courses”.

  1. create workflows capable of summarizing the product page of each competitor in a row of a table, containing information such as (where and how to create these text variables?):
  • title of the online course
  • program of the online course
  • description/presentation of the online course
  • start date of the online course
  • end date of the online course
  • timetable of the online course
  • name and surname of the teacher or teachers if there is more than one
  • price of the online course
    To do this, can I use pixelml webscraping or better use firecrawl or apify or an ai llm? Does the pixelml scraper use agenticflow credits or do I have to pay those credits separately?

  1. Create a tool that every week/month allows me to identify new online courses from competitors and extrapolate the list of new product/“online courses” urls to be analyzed by the scrapers that I have already configured in point 1.

  2. Save the data inside agenticflow, then from there can I possibly export the data manually?

2

Hi,

To proceed with your analysis of online courses, I recommend using Firecrawl API for scraping, as it will provide more control and flexibility for extracting structured data like the course title, description, timetable, etc. Pixel ML’s web scraping is more suited for specific tasks, and it may require more technical setup, so it could be a bit more complex if you’re not familiar with the details.

For your workflow:

Use Firecrawl API to scrape competitor websites and collect course details.

Set up a process to extract data such as the course title, program, description, start/end dates, teacher names, and pricing.

To save the extracted data into AgenticFlow, you can simply check the workflow logs — they automatically store the output of each run.

After storing data in AgenticFlow, you can manually export the data if needed for reporting or further processing.

For automation, you can configure the scraper to identify and extract new course URLs every week or month. If you’d like this feature to be fully built-in, feel free to suggest it on our roadmap
→ https://agenticflow.featurebase.app/roadmap

Regarding the credits, Pixel ML services are billed separately from AgenticFlow credits, so they will not be deducted from your available AgenticFlow credit pool.

:small_orange_diamond: AgenticFlow credits are mainly used to run workflows (e.g. 3 credits per step).

:small_blue_diamond: Pixel ML credits work like OpenAI or Claude credits — they are separate and used only when you call Pixel ML services (e.g. scraping, AI models). You’ll need to top up or connect a key.

AgenticFlow does not cover Pixel ML usage.

Can you drop your top 10 course page URLs now, I’ll make a short video tutorial showing how to build this template workflow for your use case.

Best,
Wendy