Calligo created a machine learning model for H-E-B to analyze data, make better inventory and pricing decisions, and suggest substitutes.
Like all retailers, leading grocery client H-E-B loses sales when products are scarce or displayed in sub-optimal positions in-store.
They asked us to design, build and manage a machine learning model that could be used for data analysis to help them make better decisions about inventory and pricing.
H-E-B also wanted the technology to suggest product substitutes that might not be immediately obvious.
Calligo developed a user-friendly Machine Learning as a Service product for H-E-B, using Databricks to analyze cart data.
Our data scientists developed a MLaaS product using an algorithm that could analyze historic cart data and make informed recommendations when products were unavailable.
With MLaaS, a business essentially hires an entire end-to-end data science team without having to recruit and retain additional experts or fund expensive new IT systems that take months (even years) to build.
But the best model in the world is useless if it doesn’t get used, so deploying it in a user-friendly way was key.
We developed H-E-B’s MLaaS product using the Databricks platform, collating the results into a visual dashboard that set out the data in a clear and easy-to-understand format.
Calligo created an MLaaS product using an algorithm that analyzed cart data and suggested product substitutes, deployed via Databricks platform.
The bespoke product provided H-E-B with a cost-effective and efficient way to analyze and make predictions from its data.
It has helped them determine what products should be placed together because they’re likely to be sold together and what products to substitute when others are scarce.
The technology has allowed for more effective use of data, increased product efficiency and maximized H-E-B’s revenue opportunities.