Smart Retail Solutions and ProfitTrax

The problem

Grocery chains sit on an enormous amount of data: sales by department and by week, shrink and known loss, out-of-stocks, item variety and assortment, and goals against actuals. ProfitTrax already turns that into reports. But getting an answer still meant knowing which report to open and how to read it.

SRS wanted a manager to be able to ask “which days last week had weak availability in my meat department, and how does my variety compare with my sister stores?” and get a correct answer. Not a plausible one.

What Rana provides

• Rana Plug-ins inside ProfitTrax: a shared prompt library, summarize and explain controls, and prompt optimization, licensed to SRS in July 2026.
• Rana Brain via the API: each grocery chain gets its own key, its own spending caps and its own usage reporting, so SRS can offer AI features to its customers with clear costs.
• Rana Dive: plain-English questions answered from each chain’s own data, inside a security model SRS controls end to end.

What we learned building it

Most of the hard work was not the model. It was vocabulary: teaching the system what a grocer means by “variety”, “known loss” or “GM”, which department names have common abbreviations, and when to ask a clarifying question instead of guessing. That glossary now lives with SRS’s own data, so it improves every answer without exposing the database.

We also learned that honesty beats fluency. When a filter is dropped or the data doesn’t support a confident answer, the system is designed to say so rather than return a number that merely looks right.