Hands-on AI application ideation workshop

TL;DR  

We recently participated in a client event where we hosted an AI ideation workshop for over 60 participants. The experience was enlightening as we guided them through the hands-on process of creating AI applications and sparking creativity to generate ideas that could enhance their work. Participants developed a wide range of applications, from job assistants to help desk improvements, and even fun applications like fishing recommendations. Once participants grasp how Large Language Models (LLMs) work and are given the tools to build applications hands-on, the potential for innovative and creative solutions becomes limitless.

About the event  

The workshop was part of the annual team event organized by the Commerce department of Transgourmet Germany. Transgourmet is one of Europe’s leading wholesalers and distributors for food and non-food products, catering to the gastronomy, hotel, and foodservice sectors. With over 30,000 employees and an annual revenue exceeding 9 billion euros, Transgourmet operates in several European countries.

Transgourmet’s websites utilize the open-source CMS Drupal with a headless commerce solution implemented as micro frontends, catering to the diverse technologies and needs of different business units across various countries. In collaboration with the software company foryouandyourcustomers, we at 1xINTERNET help Transgourmet innovate and digitalise their operations.

At this annual event, business units and partners from Central and Eastern Europe gather to reflect on the past year, discuss development activities, participate in workshops, bond with each other, and prepare the roadmap for the coming year. The event had more than 60 participants from across Europe at a conference center near the Transgourmet’s German headquarters.

Conducting the AI workshop  

Two workshops were held at the event, one of which was an AI workshop led by our CTO, Stefan Weber, and me. The participants came with varying levels of knowledge about generative AI and Large Language Models (LLMs). 

To get everyone on the same page, we started with a brief introduction, based on an AI presentation I gave at the Drupal Developer Days Conference in Burgas, Bulgaria. The main takeaway was that generative AI can be instructed with context to interact with users in specific ways.

Once the participants understood how to use this technology with data not already embedded in the model, they quickly saw its potential for their daily work. We introduced them to the strong AI integration capabilities within the Drupal framework, particularly the Drupal AI module family, and utilized our AI playground at https://ai.1xinternet.de, which we developed in 2023 for testing and demo purposes.

We divided the participants into groups of around five, ensuring each group included a mix of developers, business staff, designers, and marketers. Each group received a chat application setup where they could provide instructions, input data, and test different models and settings. This setup allowed them to brainstorm and experiment with their own chat applications. Being able to create these rapid prototypes easily sparked even more inspiration.

AI ideas generated by the participants  

The results were nothing short of extraordinary. When people are given the time and freedom to experiment and explore, they often uncover valuable and unexpected scenarios for everyday work. 

Here are some of the ideas that stood out for me:

Issue Board Assistant: One team created a chat assistant to navigate agile work environments. This is highly useful for Transgourmet because many developers and stakeholders work together, and it can be challenging to follow what is being worked on. By downloading issues and providing them as CSV context to the assistant, it could answer questions like:

•    What is currently being worked on?
•    Who is handling topic XYZ?
•    What is team member XYZ working on?
•    What is the focus of the current sprint?

Job Agent: Another team compiled a list of Transgourmet’s job offers from various portals, allowing users to search for jobs through a chat assistant. It could answer queries such as:

•    What jobs are available in field XYZ?
•    What jobs are available in location XYZ?
•    I want to work part-time; what jobs are available?

Handover Assistant: One group addressed the challenge of handovers when key personnel go on vacation. Using pre-prompted handover documents, they created an assistant that allowed colleagues to ask questions as needed without reading everything in advance.

Fishing Advisor: A fun application came from a group with expert knowledge of fishing in German lakes. They created an assistant that provided fishing spot recommendations, equipment tips, and interactive elements like images and product links for purchasing.

Participants enjoyed instructing the assistants in various creative ways. Some assistants were designed to be rude and snarky, while others were funny and sarcastic, even responding in rhymes or "youthful" language.

Stefan and I also prepared an AI assistant that could recommend products based on a large e-commerce data set of almost 30.000 orders. Using 200 orders from the dataset consisting of nearly 1,500 different products, the assistant suggested additional products to users with items in their shopping carts.

One significant benefit for Transgourmet was the multilingual capability of the chat applications. With participants from different countries, the assistants could interact in German, Romanian, Polish, Bulgarian, and more, alongside English. For a European company, this multilingual functionality is crucial.

Thoughts and recommendations for workshops to create AI solutions

AI is everywhere these days, but so much misinformation exists that it can be hard for most people to grasp how it truly works. If you want to ignite creativity and help people generate practical use cases, it’s not enough to explain the theory - you must show them and let them try it out.

Creating an engaging environment, like our AI playground and providing inspiring examples can help participants develop their own ideas.

If you'd like us to conduct a similar workshop for your team or empower you to do it yourself, let us know. We’d be thrilled to help you unlock AI's potential within your organization.

Conduct your own Ideation workshop for AI Application Ideas

Get in touch

Contact us to conduct an Ideation workshops for AI Application Ideas with your team!