Bar Digitalization - 0.0.2

Thank you very much for taking the time to read this article!

I love alcohol, and I love computers. But I can’t just remain an enthusiast anymore; it’s time for the next stage.

Link —> Personal Resume
Link —> Bar Digitalization 0.0.1

Business Operation Framework and Digitalization

Business Operation Framework

Digital Applications

In this 2023 article, several bar digitalization projects are listed. Link

I’ve placed the projects mentioned in the article into the framework above.

Table 2: Distribution of Digital Technologies in the Above Framework

A very interesting finding is that the “digitalization” of the bar counter & guest areas is quite ‘bullshit’.

When taking look at the table, I think the very things I’d love to do or optimize is happening in the “kitchen”. Or something “cocktail” oriented. It could be something like smoother work chain for bar backs, or better storage solution for recipes. And next step is that I need to collect more information about the “pain points” from the bar mans.

Conversations - Pain Points

After discussing this with a bartender & owner from my hometown (referred to as “Owner A” below), his pain points are all represented in the tables above.

Pain Point 1: Unified Quality Standards

Involves:

• Row: Internal Process Perspective
  
• Column: Bar Premises - During Business Hours, Non-Bar Premises

This issue has two dimensions: single location and chain operations.

For a single location, Owner A mentioned that different batches of fresh ingredients like lemons/limes show quality variations. For experienced bartenders, they might have the ability to flexibly adjust recipes to handle these differences, but this is more difficult for apprentices or less experienced bartenders. Could there be quality standards for this?

For chain operations, Owner A mentioned that if they opened another bar, they would want to ensure customers get the same experience with a Whiskey Sour at both locations. How? There are even more uncertain factors.

Pain Point 2: Product Waste Rate

• Row: Financial Perspective, Learning & Growth Perspective
  
• Column: Bar Premises - After Business Hours, Non-Bar Premises

Owner A described that when creating drinks, there’s quite a bit of beverage/material waste. Employee training also creates beverage waste, as well as other small things like complimentary drinks, or customers wanting a replacement for drinks they don’t like.

Currently, Owner A uses estimation to figure out the approximate cost of this beverage and material waste. He wants a more accurate solution.

My Ideas

I’m very grateful for Owner A’s enthusiastic answers, and of course for Friend B’s inspirational thoughts.

For Owner A’s pain points, I think feasible solutions would mainly involve IoT (Internet of Things). The main difficulty is at the information collection level, such as collecting quality data for lemon juice and then processing it (cross-disciplinary). Waste rate control is similar - finding an elegant way to obtain data without interfering with normal human activities. After all, it’s not possible to record every pour of alcohol in a notebook, haha!

Before talking with Owner A, my own ideas were mainly in the Learning & Growth Perspective and Internal Process Perspective from Table 2. I want to do, or rather I think the following two things would be meaningful:

• Using voice recognition/handwriting recognition + large language model semantic analysis to automatically archive drink recipe logs
• Customizing a knowledge base for the bar itself

1. Using Voice Recognition/Handwriting Recognition + Large Language Model Semantic Analysis to Automatically Archive Drink Recipe Logs

In my own home bar practice, when testing drinks, recipes often go through many iterations. Recording the proportions and flavors of each recipe often only stays in my head (because my hands aren’t clean enough to write or type at this time). When there are many trials and I get tipsy, I forget the drink recipes.

So I want to use voice recognition/handwriting recognition to get basic log “data,” for example, before making a drink, speaking into a microphone about which ingredient is how many milliliters. Then a large language model can robustly format this unstructured natural language. For writing to log files or databases, this can use an agent or traditional automation solutions.

2. Customizing a Knowledge Base for the Bar

This inspiration comes from Flavor Group’s 2023 tenth anniversary menu. I was very pleasantly surprised that modern menus can detail information to this degree. The hierarchical structure is clear, making it perfect for standardized storage!

Simply put: it’s about detailed recording of the specific milliliter amounts for each recipe of every drink, garnishes, glassware, flavor descriptions, etc. For each homemade ingredient, also detailed recording of recipes, proportions, and production techniques.

For cocktail brands with rich menus, creating a knowledge base can:

• Efficiently store the bar’s digital assets
• Facilitate information lookup and employee training

Of course, it’s not just about recording recipes; the bar can choose to put any information in this knowledge base.

A knowledge base is not the same as a database. Simply put, it’s a database plus. Imagine it as a Wikipedia dedicated to this bar, where users (customers/industry peers/apprentices) can explore by searching for entries.

Also, thanks to structured storage, the bar’s content can be easily integrated into various scenarios:

• Digital ordering
• Book publishing and content licensing
• Inventory management software data synchronization
• Employee training
• Supply chain and inventory optimization (one ingredient used in multiple cocktails)
• Creation of visual information charts
• Integration with large language models to explore unlimited possibilities
• …

Fig.1 Example of flavor group how they organize recipe data

Fig.2 The related self-made ingredients