All products

Check data freshness with AI

Find the latest timestamp in a warehouse table, compare recent row counts and check whether an apparent delay makes sense for the normal loading schedule.

Before a reporting meeting, you can check data freshness with AI to find out whether today’s numbers have actually arrived. The checking-freshness skill looks for an appropriate timestamp, measures its age and compares recent activity. A chart that appears to show a quiet morning may simply be waiting for its next load.

The timestamp you choose matters. The date on an invoice can be weeks older than the time it entered the warehouse. The report makes that choice visible and can include the producing Airflow workflow when a stale table needs investigation.

What it does

  • Identify a load or update timestamp in a known table.
  • Measure the time since the latest value in hours or minutes.
  • Compare daily volumes over the last week to spot missing loads.
  • Interpret a delay against a daily, hourly or occasional schedule.
  • Check whether an authorised Airflow pipeline is paused or failed recently.

Install and use

  1. Install the skill and provide a SQL client with read access to the warehouse. For pipeline follow-up, use an authorised Airflow UI or the relevant first-party CLI.
  2. Ask, “Can I use the refunds table for this morning’s review? Check when rows arrived and compare the past seven days.”
  3. Read the timestamp definition and expected schedule alongside the status before using the table.

FAQ

Source

Made by Astronomer. Original skill: checking-freshness, published on GitHub under the Apache-2.0 licence.

View on GitHub