3 ms·

Any more details about how it was built?

• vinothgopi 11y ago
  Geckoboard part was relatively easy - they have a nice interface to design beautiful looking dashboards. Main magic was in the data pipeline. We tagged products that are Black Friday specific as seen on the merchant specific "Black Friday" category pages and other aggregators. These sites' pages were checked and the products' prices updated more frequently than our regular pipeline. Everything ran as part of the Semantics3 infrastructure but at the last step we piped the updated information of the Black Friday specific products to a separate Postgres DB as well and that allowed us to run fancy queries that could answer almost any kind of questions we have about the data.
  • themckman 11y ago
    I'm interested in knowing more about the Perl component. What can you say about that?
    • vinothgopi 11y ago
      Our entire data processing pipeline is actually written in perl (that is several 10s of thousands of lines of code). So is the subcomponent that actually pipes the data to Geckoboard.