{"id":4515,"date":"2026-03-12T19:18:57","date_gmt":"2026-03-12T17:18:57","guid":{"rendered":"https:\/\/agentaya.com\/?p=4515"},"modified":"2026-03-13T18:17:06","modified_gmt":"2026-03-13T16:17:06","slug":"how-does-ai-fit-in-with-traditional-whisky-making","status":"publish","type":"post","link":"https:\/\/agentaya.com\/fr\/how-does-ai-fit-in-with-traditional-whisky-making\/","title":{"rendered":"How does AI fit in with traditional whisky making?"},"content":{"rendered":"\n

This post was originally published on manxwhisky.com<\/a> by the Manx Whisky Company , a small, traditional distillery based in Baldrine on the Isle of Man. Founded in 2019 by Magnus Grinneback and Scott Young, it focuses on producing authentic single malt whisky using locally sourced materials. We are publishing it on AgentAya<\/a> with their permission.<\/em> You can find the original article here<\/a>.<\/em><\/p>\n\n\n

\n
\"manx<\/figure>\n<\/div>\n\n\n

Our whisky-making is very traditional. We collect barley from Ballavell Farm, and much of what follows would have been familiar to a distiller 250 years ago (you can read more about our setup here<\/a>).<\/p>\n\n\n\n

So, how is Artificial Intelligence fitting in with this?<\/p>\n\n\n\n

In quite a few ways, actually. We pay close attention to developments in AI and constantly consider how it can make our work easier. Let\u2019s look at a few examples – from simple to more advanced. <\/p>\n\n\n\n

<\/span>The Basics: How we are using AI for day-to-day tasks<\/strong><\/span><\/h2>\n\n\n\n

1. Copywriting – tightening up communications<\/h3>\n\n\n\n

We don\u2019t use AI to write<\/a> from scratch. Instagram posts, blog posts (including this one) and newsletters all start with us writing a draft.<\/p>\n\n\n\n

However, we do use AI to polish initial text. This speeds up the journey from first draft to finished post. We’ve experimented with training a GPT to write posts from scratch, but the results aren\u2019t good enough – too many clich\u00e9s and too much sensationalism. For now, we prefer to write in our own voice and use AI to refine rather than replace it.<\/p>\n\n\n\n

2. Small coding jobs – improving customer notifications<\/h3>\n\n\n\n

Our website runs on Shopify, which is a solid platform. Occasionally, we need to adjust the logic – for example, sending different confirmation emails depending on what a customer has purchased. A tasting booking requires very different information from a bottle purchase.<\/p>\n\n\n\n

Here AI has been extremely useful. While I have some coding experience from my student days 30 years ago, I\u2019m completely lost when it comes to Shopify\u2019s scripting environment. AI allows us to make these changes ourselves rather than hiring a developer.<\/p>\n\n\n\n

Looking ahead, we may replace some of the paid modules we currently use (age verification, for example) with our own AI-assisted code. These modules aren\u2019t expensive, but if we can save \u00a310 per month and build something better suited to our needs, it\u2019s worth doing.<\/p>\n\n\n\n

3. Admin automation – simplifying monthly distillery returns<\/h3>\n\n\n\n

Running a distillery involves monthly declarations of production and warehouse withdrawals. That means forms – not the most exciting part of the job.<\/p>\n\n\n\n

We\u2019ve used AI (Claude, in this case) to write a script that automatically populates the forms based on our production data. It\u2019s hardly revolutionary, but it saves 5\u201310 minutes each month. More importantly, it has shown us how easy it is to build small scripts that automate specific tasks<\/a>.<\/p>\n\n\n\n

\"manx<\/figure>\n\n\n\n

<\/span>How we are using AI to build distillery tools<\/strong><\/span><\/h2>\n\n\n\n

1. Building our own measurement app<\/h3>\n\n\n\n

In a typical distilling week, we perform numerous measurements and calculations: mash water temperatures, temperature corrections for alcohol and sugar readings, dilution calculations, and more.<\/p>\n\n\n\n

There are plenty of online tools that handle these individually. However, they rarely work exactly the way we do, which often means hopping between websites and manually stitching together results.<\/p>\n\n\n\n

To solve this, we’ve built our own app tailored to our workflow. It contains everything we need in one place. This simplifies work in the distillery, reduces time spent cross-checking figures, and – crucially – lowers the risk of mistakes.<\/p>\n\n\n

\n
\"\"<\/figure>\n<\/div>\n\n\n

This is an example of how we’ve tailored a standard mash calculator to our particular setup. Without going into too much detail, it takes preheated water into account and calculates both the protein rest we always perform and the required strike water temperature.<\/em><\/p>\n\n\n

\n
\"\"<\/figure>\n<\/div>\n\n\n

Here is another example. This is a standard hydrometer temperature correction tool, but it is linked directly to the original gravity of the fermenter. There is no need to perform a separate temperature correction. Everything is calculated in one place, giving us the alcohol reading and attenuation without jumping between screens or tools.<\/em><\/p>\n\n\n\n

\"cask<\/figure>\n\n\n\n

2. Cask and distillery management software<\/h3>\n\n\n\n

Until recently, we managed our casks and inventory using Excel. It worked – but it was clunky and time-consuming. Making mistakes is easy, and tracking which mash went into which cask requires full focus.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

This was an example of our Excel-based cask overview. In addition to this, each cask had its own individual worksheet. As our inventory grew, this quickly became difficult to manage.<\/em><\/p>\n\n\n\n

We\u2019ve now used AI to help build our own<\/a> cask management software. With a click, we can see the full history of each cask, its age, and related production data. The system also reminds us when sampling or other actions are due.<\/p>\n\n\n\n

\"overdue<\/figure>\n\n\n\n

Above you can see a screenshot showing overdue actions and a list of our active fills – quite a contrast to the Excel-based solution.<\/em><\/p>\n\n\n\n

We are currently working on integrating mash and distillation records into the system. The final step will be to add scripts that automatically prepare our monthly production and withdrawal declarations for Customs & Excise.<\/p>\n\n\n\n

\"mash<\/figure>\n\n\n\n

Here you can see part of our previous mash log – many manual entries and a significant amount of calculation.<\/em><\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

In contrast, here is the new system – clean, structured, and much easier to navigate. It\u2019s still new, but we expect it to significantly simplify record-keeping – a critical part of running a distillery.<\/em><\/p>\n\n\n\n

<\/span>Using AI for research and distillery design<\/strong><\/span><\/h2>\n\n\n\n

AI is also an excellent research assistant<\/a>. If we\u2019re curious about something – for example, how high-protein barley might affect flavour – AI can give us a structured introduction in minutes. That helps us get up to speed quickly before deciding whether to explore further.<\/p>\n\n\n\n

It\u2019s also useful when planning physical improvements. We are currently working on an improved hot condenser water recovery system. AI has helped us design the electronic control system for the pump, suggest components, estimate costs, and identify suppliers.<\/p>\n\n\n\n

Much of this could be done without AI but it would take significantly longer.<\/p>\n\n\n\n

\"manx<\/figure>\n\n\n\n

<\/span>Final thoughts<\/strong><\/span><\/h2>\n\n\n\n

AI is allowing us to research, design and build systems that would previously have been too costly or time-consuming for a small distillery. The tools themselves are inexpensive. We use both ChatGPT and Claude. Each costs about 20 USD per month for a pro subscription.<\/p>\n\n\n\n

AI isn\u2019t going to change how we malt, mash, distil, or mature our whisky. But as a small company, we find it to be a powerful companion – one that reduces time spent on administration and increases time spent where it matters most: making whisky and improving our production.<\/p>\n\n\n