Method
Where each story comes from and what it can and cannot tell you.
Vendor case studies
Vendor case studyCustomer-story pages that Snowflake, Databricks, MongoDB, Confluent, dbt Labs, Fivetran and others publish, found through their sitemaps each week and read as text.
Customer blogs and keynotes
Customer blogA company's own engineering blog or a conference talk it gave, when it names the vendor and the use.
Press releases and news
Press releaseAnnounced partnerships, renewals and extensions, found with a few news searches a week and read from the article itself.
Public contracts
Public contractUS federal awards on USAspending.gov whose description names the product: the agency, the amount, the dates. An award that says renewal, or repeats, counts as a renewal.
How a story is kept
Each story is one customer, one vendor and one source: the products named, a one-line use case, any outcome the source itself states, the relationship (adopted, expanded, renewed, moved to, consolidated) and the date the source gives. A migration also records the platform the customer left, which shows as a loss for that vendor.
New finds come from a weekly run, Saturday at 10:00 UTC. Pages are read as text and a model turns each into the fields above; a story is kept only when the sentence supporting it is found on the page, and it shows Unverified until checked by hand. Federal awards are read straight from the USAspending.gov API.
From public case studies, filings and announcements. Vendors publish their own success stories; check the source. A case study shows that a company used a product for a purpose when it was published; it says nothing about spend, and the company may have moved on since.
Questions
What is Stories?
Stories is a database of how named companies use data, cloud and AI vendors, built only from public sources: vendor case studies, customer engineering blogs, keynote talks, press releases, earnings calls and public contract records. Every story links to where it was published.
How do I see a company's tech stack?
Open the company, for example Medtronic, and its page groups every vendor it is publicly known to use by category (cloud, data cloud, lakehouse, streaming, transformation), with the use case and the source for each. Vendors it moved away from are listed separately.
Where do renewals and losses come from?
Renewals come from announced partnership extensions and from public contract records such as USAspending.gov, where an award for the same product repeats or says renewal. Losses are stories in which the customer says it moved from one vendor to another, usually in the new vendor's case study.
Are vendor case studies reliable?
They are the vendor's own marketing, approved by the customer. They show that a company used a product for a purpose at the time of publication, not that it still does or how much it spends. Check the date and the source.
How often is it updated?
A run every Saturday at 10:00 UTC reads new pages from vendor customer-story sitemaps, new federal contract awards and a few news searches for renewals and migrations. New finds show an Unverified badge until they are checked.
Which companies moved from Teradata or Hadoop to Snowflake or Databricks?
The Migrations page lists every story in which a customer says it moved from one platform to another, with the source: for example from Teradata, Cloudera or on-premises Hadoop to a cloud data platform.