AI-powered demand forecasting for grocery

Guac accurately forecasts demand for grocery retailers, to reduce food waste and increase availability.

Jobs at Guac

New York, NY, US
$130K - $200K
0.50% - 1.20%
1+ years
Team Size:4
Location:New York
Group Partner:Tom Blomfield

Active Founders

Euro Wang

Co-Founder & CEO of Guac. Passionate about tackling food insecurity & waste with technology!

Euro Wang
Euro Wang

Jack Solomon

Co-Founder and CTO of Guac. ML engineer working to make grocery retail more efficient.

Jack Solomon
Jack Solomon

Company Launches

TLDR: We tell grocery stores exactly how many avocados they’ll sell next Tuesday.

Guac uses machine learning to accurately predict how much grocery retailers will sell each day. This helps them order the perfect amount of inventory: enough to meet customer demand without sending tons of leftover food to landfill.


  • Predicting future demand is hard: product-level demand drastically fluctuates each day based on hundreds of factors like weather, road closures, and sports games. Yet, even the largest supermarket chains still use basic regression models or Excel formulas, which means …
  • Tons of food waste and empty shelves: which costs grocery retailers in Europe/US $188B and costs our planet 15 million metric tons of greenhouse gas emissions each year.


  • Industry-leading algorithm that predicts order quantities for each item, each day, each store location. Our average forecast error (MAE) for our current customers is 0.95 units, which translates to a 38% reduction in food waste.

  • How?
    • More than just ML: Grocery demand is driven by real world events that ML models don’t know exist. So we inject 230+ external variables to actually capture people’s buying behavior …
    • For example, we found that sports betting odds help us predict beer sales. When a sports match is predicted to be a close game, it turns out more people watch the game, driving up beer sales!
    • Customized for each store location: grocery stores located just 5 miles apart have different customer behavior.
  • Inventory ordering optimization: We then turn our hyper-accurate predictions into useful inventory order quantities, based on individual SKU-specific requirements: minimum order quantities, promotions, unknown loss estimates, shelf life, supplier lead times, and more.
  • Modular integration: We integrate our ordering recommendations right into the retailer’s existing inventory ordering softwares and workflow. Whether inventory ordering is done at the HQ-level vs. store-level, we fit into the retailer’s system, instead of forcing them to uproot their system to fit into ours.


We’d love introductions to grocery chains (anywhere in the world), especially:

  1. Discount grocery chains, with tighter margins
  2. Grocery chains in more remote locations (e.g. Hawaii), with longer supplier lead times and inflexible supply chain
  3. Asian grocery chains in the US, who import a lot of their products from abroad

And if you know anyone in the general food/grocery space (POS providers, suppliers/distributors, grocery food waste charities, etc), please let us know at founders@tryguac.co.