What are the differences between Kadoa Monitors, Pipelines, and Datasets?

Updated

Kadoa groups its web data offering into three products with different jobs: Monitors alert you when something changes on the web, Pipelines automate end-to-end scraping and keep it running as sites change, and Datasets are curated collections built around a specific investment thesis from public web data. Choose Monitors when you need to know when something happens, Pipelines when you need a repeatable feed of data you define, and Datasets when you want a ready-made dataset tied to a research question—with a free sample available to test before you pay.

Side-by-side comparison

Dimension Monitors Pipelines Datasets
Primary job Catch market-moving events first Automate web scraping end to end Deliver a dataset built around your thesis
What you get Alerts on filings, prices, and website changes A maintained data flow you define A curated dataset from public web data
Who does the setup You point at sources; Kadoa runs the monitoring You describe the data; Kadoa agents build and run the pipeline Kadoa builds it around your thesis
Maintenance Handled by Kadoa Scraper creation and repair automated as websites change Built and maintained by Kadoa
Best fit Event-driven and time-sensitive research Recurring data needs that must survive site changes One-off or thesis-specific research where you want a finished dataset
Try before committing Not specified in available materials Not specified in available materials Free sample before you pay

Monitors: fastest alerts on changes

Monitors are positioned as the fastest, most reliable way to be alerted to filings, prices, and website changes. The value is timing: you find out about a market-moving event without polling sources yourself. This fits workflows where the trigger matters more than the full dataset—regulatory filings, price moves, or a page quietly changing.

Pipelines: automated, self-maintaining scraping

Pipelines take a description of the data you need and have Kadoa agents build, monitor, and repair the scraper. The stated design goal is that data keeps flowing as websites change, rather than breaking within a week and requiring manual QA. Kadoa says 85% of incidents are resolved automatically and that the time from request to first dataset is under two hours. This is the option for recurring needs where reliability over time is the main cost driver.

Datasets: built around your thesis

Datasets are assembled from public web data around a specific investment thesis. Instead of defining a scraper, you get a finished dataset, and Kadoa lets you test a free sample before paying. This suits research where the question is well-defined and you want the data delivered rather than the pipeline managed.

How they connect

All three sit on one data layer and can deliver into tools teams already use: Excel, Google Sheets, and Slack alerts for analysts and PMs; Snowflake, Databricks, BigQuery, S3, SFTP, webhooks, and API for engineers; and MCP, Claude, ChatGPT, and internal agents for AI workflows. In practice, a Monitor can flag an event, a Pipeline can keep the underlying feed current, and a Dataset can package the result for a specific thesis.

How to decide

  • Need to react to events quickly? Start with Monitors.
  • Need the same data refreshed continuously despite site changes? Use Pipelines.
  • Need a finished dataset for a defined research question? Request a Dataset and test the free sample first.
  • Unsure? The three are complementary, so the choice is about whether your bottleneck is timing, maintenance, or assembly.