One small step for Semantics3, one long-legged leap for product data
Five years ago, we set out on a mission to organize the world’s retail information. While we’ve been proud of the progress made so far, there’s been an elephant in the room that’s haunted us since the very beginning — our focus has been limited to ecommerce data alone. The future of retail, big box retail outlets, has been beyond our reach.
Today, we’re setting this wrong right. We’re pleased to announce, that for the very first time, Semantics3 will be offering real-time in-store catalog prices to power the next generation of commerce intelligence.
Challenges in indexing physical retail data
There are a number of hard problems that had to be solved for us to be able to deliver on the promise of high quality in-store information.
Take these two issues for example,
- Market coverage: We had to make sure that we had a fully representative sample of all products being sold. But data from mom and pop shops in rural areas can be difficult to physically access and transmit.
- Refresh rates: We had to keep availability & pricing information up-to-date. But store managers have been known to alter prices at whim, without notice.
We knew that these problems could not be solved using traditional web crawlers. We needed to go above and beyond. We needed something that could act as a single-shot solution to all of the complexities at hand. We needed … AI!
Semantics3’s new AI team
Without further delay, I’d like to introduce the new stars of our Arachnid Intelligence team — our very own cluster of spiders, all set to crawl their way into retail stores all around the world.
Muffet, leads our cluster of mouse spiders, Littlius missulena. Specializing in indie and small retail stores, this team works on making sure that the mom-and-pop store around the corner never feels left out.
Next up, we have the largest division within our AI team, long-legs, Shelobius tolkenius. Led by Samwise, you are most likely to encounter them diligently covering the Whole Foods and Target stores in your city.
Finally, we have our heavy-lifting tarantulas, Aaragopelma forbidi. Rubeus and his aides specialize in cataloging the warehouses of the world. At ease in high-roofed storage depots, no Walmart is too big and no crate too heavy when it comes to making sure offline retail information gets online.
Trained at our specially built AWS (Arachnid Web System) gym, using reinforcement learning techniques, our AI team is all set to achieve state-of-the-art performance (in both precision and recall) in mining and parsing product data.
Get your invite today!
Starting today, we will be rolling out access to the physical retail data in batches. If you can’t wait and would like access now, email us at firstname.lastname@example.org and we’ll hook you up.