How Åben’s operations team became AI savvy

About Åben

ÅBEN is a Danish craft brewery founded in 2017 by Philip Hulgaard initially operating from a family farm in Kolding.

The company has since established its brewery and headquarters in Copenhagen’s Kødbyen, combining beer production with a public bar and hospitality experience.

ÅBEN operates 6 bars across Denmark, sells beer online, employs 90 people and is involved in festivals including O Days and Kalas.

Recognised as Denmark’s Brewery of the Year in 2019, the brand is known for quality-focused, experimental brewing and a strong social identity.

Challenge

The Åben leadership saw the opportunity presented by AI to help improve the operations throughout the business. Their ambition was to quickly arm their team with the powerful new tools and enable them to optimize and automate their work. On the other hand, the team naturally worried about security and exposing certain data or systems.

Åben does not have an internal IT team. And when we first met the Åben team, the general experience with AI was summed up to “asking Chat”. While most had seen some value, few had had a real “aha” moment.

The goal was to bring Åben’s team and organisation to a place where they benefit from and use AI tools naturally to improve their work.

Approach

We started with a simple plan:

1. Define a core team as a beachhead for the rollout
2. Define and implement the Claude setup and guardrails
3. Train the core team
4. Help the core team build solutions and automations to stack initial wins
5. Iterate on our approach with learnings
6. Continue the rollout to other teams

So first we set the team. Johan Thybo, the Head of Finance, became the AI responsible and internal project owner to run the project on Åben’s end. He was joined by 4 department leads who knew the internal processes well and who would later help support their own teams.

Together with Johan, we set up the Claude organisation and defined the guardrails and risks. Which permissions, the data access, and what systems to connect and which not to.

We then onboarded the team and ran the training through a session of weekly practical workshops. Each started with a walkthrough of key concepts like Skills, Agents, etc. The majority of the time was spent “hackathon-style” with each of the team members working in parallel to apply AI to their own tasks and work. We helped each identify and map specific work ripe for AI application and automation. Each session ended with demos where each person showed their work. Already during the first session each person had experienced their “Aha” moment and built their first solution or automated their first piece of work.

After five weeks the core team had gained practical experience, successfully automated multiple workflows and even built their first app.

Results

In just five weeks, the Åben team went from sporadic individual AI usage to having established their organisational approach and a core team that had built automations and solutions that solve real work. Now the team has the confidence and experience to continue the roll out throughout the organisation.

Specifically, with our support, here’s what the team built:

Johan, Head of Finance, automated two recurring tasks:

1. An automatic email router that categorises incoming emails and forwards invoices directly to their payment system. This task previously required manual triage every day.

2. A spreadsheet reconciliation tool that checks two spreadsheets every three days for errors and flags them for the team. With no automation these checks can take hours each week.

Alex, Head of Bars, built an internal loyalty card management app

Åben offers friends and family different discount rates. Earlier this was easy to manage but with growth has become a repeated headache to manage. Without prior coding experience, but with our help, Alex managed to build a simple app that gives the team a complete overview of discounts and the user has a clean way to show their discount code.

Christina, Head of Logistics, automated two separate workflows:

1. A route-planning optimization tool built on Google-Maps that optimizes transportation time between locations as Åben's drivers move through the city for biweekly maintenance checks.

2. A tool to forecast required packaging volumes. Christina built an app-script that pulls data from their CRM and on a weekly basis automatically summarizes and updates the amount of packaging needed for their production volumes to make ordering easier or the logistics team.

3. An automatic email classifier and router for a shared Gmail inbox. Christina built a router that automatically reads, categorises and routes emails to the right people in the team reducing the time each person spends in the inbox and on manual categorisation.

Clara, Head of Design, developed Åben's full design system and implemented in Claude Design.

Design is incredibly important at Åben and a key part of their identity and driver of success. But the design system was not accessible outside the design team. So Clara built out the design system of fonts, colours, spacing rules and brand styling. And she ported it to Claude Design where the wider team can access it and use it to build internal presentations, social media posts, and other assets, all in the Åben style without the design team’s involvement.

Learnings

Just get started on your enablement journey

Our experience with the Åben team shows how far a non-technical team can come in a short time by taking action and getting the right support. It's clear to us that organisations should enable their teams and they too will be surprised by the solutions and creativity

Non-technical people can build sophisticated solutions today

The internal loyalty app is an example of an app that would have never been built before generative AI changed programming. The app is a quality of life improvement, not mission critical. Åben would have never hired an external partner to build it. And an internal IT team would have been busy with higher priority work. The fact that Alex, the bar manager, with a limited amount of support, could solve this own problem makes the difference between "never" and "done".

Using AI gives people agency and can be fun

The four people in the core team quickly found a tool that helped them solve problems they couldn't before. It's a form of leverage that feels empowering. Now knowledge workers can in fact change "they way we always did it".