Mini Kanban

Inspiration

The inspiration for Mini Kanban came from the need for a lightweight yet powerful project management solution for Confluence users. While Jira offers robust functionality, it often feels heavy for small projects or quick collaborations. On the other hand, Trello requires a platform switch, which disrupts workflow when the project is documented within Confluence pages. Mini Kanban bridges this gap by embedding an AI-powered Kanban board directly into Confluence pages, enabling project tracking without leaving the documentation space.

What it does

Mini Kanban extends Confluence pages to interactive project management pages with fully integrated Kanban boards. With the Mini Kanban macro, you can create a board at the touch of a button that integrates into the content of your page. Thanks to AI, users can automatically generate tasks and workflows from a user query and the existing page content.

You can efficiently manage tasks, assign specific tasks to team members and track progress - all within Confluence, without having to switch to external platforms such as Jira or Trello. Mini Kanban was developed specifically for small, page-orientated projects, but still offers all the benefits of a Kanban system. The combination of simplicity and flexibility makes it a solid solution for teams who want to organize their projects directly in the Confluence environment.

How we built it

The development of Mini Kanban started with the implementation of a simple Kanban board with intuitive drag-and-drop functionality. We used the React library DnD-kit to ensure a smooth user experience.

Next, we extended the application to allow the configuration of the Kanban board to be saved directly via the macro config. After revisiting the project for the Codegeist Hackathon, we then developed the ability to assign users to individual Kanban items to facilitate team collaboration.

Next, we integrated an AI-powered feature to automatically create a fully customized Kanban board based on a user prompt and the content of the Confluence page.

Finally, we made the DnD and AI functionality more robust, fixed UI bugs and got the app ready to release.

Challenges we ran into

There were a number of technical challenges in developing Mini Kanban. One of the main difficulties was keeping the height of the macro flexible. Each time a new item was added, the height of the iFrame would have been adjusted, which in turn would have unintentionally triggered dnd-kit's drag-and-drop mechanism, as it interprets a change in height as 'releasing' the item. After much research and analysis of possible workarounds, we decided to keep the height static to ensure a stable user experience.

Another challenge was storing the board configuration. Our goal was to implement a solution that was as simple and robust as possible. Options such as page properties proved to be too complex and inappropriate, as the configuration could be lost when a page was exported. We therefore decided to store the configuration directly as a macro config parameter. This solution is not only simple, but also preserves the integrity of the configuration during page operations.

What we learned

We learned important lessons during the development of Mini Kanban. One key lesson was that it is often better to choose a simple and pragmatic solution than to design an overly complex architecture. This was particularly evident when it came to storing the configuration, where a simple implementation using macro config parameters was the most efficient and reliable choice.

We also learned a lot about how best to design AI prompts to get accurate and useful results. This knowledge was crucial in refining the functionality to automatically generate Kanban boards based on page content and user input, and to further improve the user experience.

What's next for Mini Kanban

Next, we want to implement the integration of Confluence tasks into Mini-Kanban. The aim is to synchronise each item in the Kanban board with a corresponding Confluence task. This will allow users to manage tasks directly from the board and benefit from a connection between the Kanban workflow and the native task management features of Confluence. This enhancement will further improve collaboration and traceability within teams. We are also looking to replace the OpenAI API with Atlassian Rovo. By replacing the OpenAI API with Atlassian Rovo, we can leverage its integration with the Atlassian ecosystem, AI-powered assistance, and enhanced task management capabilities to further improve collaboration and traceability within teams.