Food & Beverage · Spirits2025Product build
BourNosh

An AI chat app that pairs bourbon with food

Wine has Vivino and beer has Untappd. BourNosh set out to give bourbon its own guide: a chat that pairs a bottle with food, food with a bottle, and suggests what to try next. We built the app and its bourbon database. It did not launch.

BourNosh
10,000+
bourbons collected with flavor data
3
ways to pair: bourbon, food, snack
3
parts we delivered: design, app and data
  • Food & Beverage · Spirits
  • Product build
  • 2025
The challenge

Overwhelming choice and zero guidance

Bourbon drinkers face thousands of bottles and little help choosing. Food, mood and occasion are ignored when people hunt for a new bourbon, even though bourbon is tied closely to food.

The founders wanted an app that talks like a knowledgeable friend: tell it what you are drinking or cooking, and it tells you what goes with it.

BourNosh loading screen, welcome message and first question
Onboarding: the welcome message offers three ways in, then the chat takes over.
What we built

One chat, three ways to pair

  1. 01

    Food to bourbon

    Tell BourNosh what you are sipping, and it recommends the food or snacks that suit it.

  2. 02

    Bourbon to food

    Describe the meal you are making, and it suggests the bourbon to pour.

  3. 03

    Bourbon to bourbon

    Name a bottle you like, and it suggests new ones with a similar flavor profile, with proof, size and why you will like each.

Read more

The chat asks short follow-up questions, such as the occasion or a flavor preference, and uses the answers to shape the pairing. Results come in short, illustrated sets, for example salty, sweet and crunchy snacks for one bottle.

BourNosh chat recommending salty and sweet snacks for a bourbon
Food pairing: the chat asks about the occasion, then answers in illustrated sets of snacks.
The data

A bourbon database with flavor metadata

Pairing only works if the app knows each bottle. We collected more than 10,000 bourbons with detailed flavor metadata: distillery, proof, size and tasting notes such as caramel, dark fruit, spice or oak.

Read more

An AI pairing engine combines those flavor profiles to rank matches. We designed the app around the key user journeys: pairing, discovery and a tasting journal.

BourNosh suggesting three bourbons similar to a chosen bottle
Bourbon to bourbon: three suggestions for a chosen bottle, each with its flavor tags.
The outcome

Built, but not launched

We designed the app, built it and collected its database. The product did not go to market.

The bourbon world is familiar ground for us: for Roth River we built IoT apps that monitor whiskey aging in the barrel.

Have a product idea with a data problem inside?

If your app depends on a catalog that does not exist yet, we can collect and structure the data, build the AI on top of it and ship the first version.

Your story next

Tell us the outcome.
We will write your case study

Discovery in two weeks. ROI proposal after. Public case study optional, by your choice.