Most SMBs make pricing decisions using intuition and inefficient methods (the old school way), yielding sub-par sales. However, with recent advances in machine learning, such decisions can now be made optimally using data to maximize sales. 

Problem and Solution

One of the most important decisions a business owner makes is setting the right prices for their products. Through our primary market research (PMR), we found that most business owners are commonly asking questions such as: (i) How much are customers willing to pay? (ii) When should I increase prices? (iii) Is it a good idea to discount prices and get the inventory moving? However, current solutions in the market are inefficient. 74% of Ecommerce SMBs in the US use the most basic method of pricing: cost-plus pricing- simply charging consumers the cost of goods sold plus a predetermined amount of profit, embodying the basic “outdated” idea behind doing business. While familiar and easy to implement, these prices bear no relation to the amount that consumers are willing to pay. As a result, profits are left on the table daily.

To this end, the main problem we solve is allowing businesses to quickly react to market forces and set optimal prices for their products, leading to more efficient sales. In essence, we try different prices to learn patterns and correlations between different external factors (inventory levels, ad spend, seasonality, location, weather, time of day, etc) and revenue. We use mathematical models to minimize price iterations while maximizing profit to arrive at the optimal price. 

If you're interested in collaborating, please reach out to:

Vision bringing machine-learning based efficient pricing to worldwide eCommerce.


Our inspiration started as a problem, grew into curiosity and culminated into catalan. We wanted to eat at our favorite restaurant, but it was too expensive. We questioned why they couldn’t reduce prices during down times. Figured out they and most others didn’t have the know-how to price. We did. 

State of the Project

Catalan is currently running 4 experiments, automating prices for Shopify stores in different industry verticals. In their first experiment, they increased the store’s average order value by 8.7% and profit by 7%!


Ishaan Grover (MAS PhD ‘23) and Andres Garza (MIT Sloan ‘22) met one year ago, in Fall 2021, while taking the Media Lab’s: MAS.665 Global Ventures: AI for Impact. Taught by Prof. Sandy Pentland and Prof. Ramesh Raskar, the class inspired them to think big with the class project prompt being: “come up with a business idea that in 10 years will reach 1 billion users and achieve $1 trillion in revenue”. Galvanized, Ishaan and Andres conducted initial PMR with stores on Central Square and identified a huge and untapped problem for SMBs: suboptimal pricing methods. During IAP, they participated in MIT Fuse where they identified the beachhead market (Shopify SMBs that make between $1M-$10M in yearly revenue). In the Spring, they took Prof. Paul Cheek’s 15.378 Venture Creation Tactics, where they commenced their first experiment with a Shopify store: This summer, they participated in MIT Delta V along with Divij Lankalapalli (MIT Math / Computer Science ‘24) where they built out the tech stack, designed and launched the consumer-facing web app, and onboarded 4 more stores to conduct experiments.