Mathematical Algorithm Determines Your Coffee Preferences


Answer a quick question: what’s your favorite coffee variety? OK, not too tricky, but can you tell me  why  that’s your favorite style? A little tougher isn’t it?

Many of us have a pretty good idea of what tastes good and what doesn’t, but matching our particular preferences to the corresponding coffee notes can often be difficult, and forget about learning the science behind our tastes ”ain’t nobody got time for that. Well, maybe one company does. Meet Bean Genius, a Los Angeles-based start-up offering a subscription-coffee service unlike anything else currently on the market.

No, silly, not this Bean Genius...
No, silly, not this Bean Genius…

The genius behind this bean reaches quite a ways past its catalog of readily available fresh roasted coffees, though the company does have a large portfolio of highly rated roasters on board. The real key is Bean Genius’s ability to isolate each customer’s exact taste preferences and match them with coffee that will perfectly satiate their individualized pallets.

Bean Genius does this by referencing a specially designed algorithm that cross checks customer satisfaction against a 30+ point rating and flavor scale.

œWe’ve broken down the coffee on our side into over 30 different flavor characteristics,  says Krishna Gupta, Bean Genius’s co-founder. œOver time, we can start to see a pattern [with what customers like or don’t like]. We analyze all these coffees. We have cuppers breaking it down and feeding data into the algorithm. 

If the business end of the organization seems a bit cold and calculating, well, that’s balanced by the bevy of creative and talented roasters the company is working with. Having already partnered with well-respected brands such as Klatch Coffee, and with more roasters coming on board all the time, Krishna feels the combination of mathematics and craftsmanship will lead to a series of highly memorable coffee drinking experiences. œEveryone’s coffee journey should be individualized,  Krishna says.

Award-winning Klatch Coffee is one of the high-quality roasters whose coffee will be offered by Bean Genius.
Award-winning Klatch Coffee is one of the high-quality roasters whose coffee will be offered by Bean Genius.

With a tentative early December launch, a fresh new website in the works, and a built-in customer base thanks to a previously successful beta test, Krishna and his partner, Ranga Yarlagadda are confident their idea will catch on with people seeking not only an excellent cup of coffee, but a better understanding of why that cup does indeed taste so great, specifically to him or her.

Each time a customer receives his box, he is asked to log onto his account and simply say whether he liked or disliked the coffee. This data, once entered into the algorithm, will be able to further analyze each customer’s preferences.

In order to make sure that each delivery is as fresh as possible, Bean Genius has worked out a system whereby the roasters themselves will be the ones sending out each order, eliminating the need for a physical middle man.

œThat was a huge lesson in our beta test,  says Krishna. œWe were sending out multiple coffee batches once a month. We found out [that] to provide freshness and to customize each box, it became a logistical nightmare. Now we’re placing orders in real time and the roasters are drop shipping for us. We are ultimately a marketplace for them to sell through. 



Jeremy Martin

Jeremy  Martin  is a freelance writer and photographer who has reported on coffee, craft beer, college sports, and business for a variety of publications over the past six years.  A veteran of the café industry and graduate of Western Michigan University, Jeremy lives in Seattle where can often be found making sandwiches from whatever is left in the fridge and cracking wise for the amusement of his adoring wife Amanda.

About baristamagazine 2104 Articles
Barista Magazine is the leading trade magazine in the world for the professional coffee community.


  1. Hi Norman!

    Krishna here, thanks for your insightful comment. You are absolutely right, our current approach is not the ‘end all/be all’ solution and your other variables are things we are considering and realize are important. I believe our solution is a better proposition than anything else out there, so I wanted to get it out in some form rather than wait.

    We are actively looking to integrate even more variables in to the logarithm outside of flavor notes to further fine-tune the recommendations.

    It’s an exciting time, we launch this week! If you, or anyone, have further questions, please do not hesitate to reach out here, or on Twitter – @BeanGenius.

    Happy Thanksgiving everyone, make sure to serve amazing coffee!

  2. Interesting!
    But how do they ´eliminate ´ the human factor and other variables?
    like machines, brewing methods, water quality, (amateur) barista skills, (not) adding milk/sugar etc.
    This “The customer simply say whether he liked or disliked the coffee.” might not be that simple as it sounds….

Comments are closed.