Current State of Development (August 24th 2018) — scaled agile development.

This blog series shows what we are currently working at. It will give you some insights on specific questions we are dealing with right now and also on our overall progress.

Last time we gave some insights on the Unibright Connector, this time we will have a closer look at our project management setup and our scaled agile development approach.

After we reached our softcap, we directly started with setting up our development team. In a blog post in May we explained:

We are working with 4 dedicated teams on the 4 different parts of the Unibright Framework (Workflow Designer, Lifecycle Manager, Connector and Explorer). We decided for a scaled agile development structure, working in scrum based 2-week-sprints. Basically this means to set up a backlog (think of this as a big to-do-list) for each of the development projects and pick items out of that backlog to be done within the next 2 weeks. After that a new sprint starts, and so on.

Now, a little more than three months later, we have the experience of about 8 sprints for each team. While in the beginning there was enough work for each team to work “in their own field”, synching between the teams now gets more and more important, as we are getting closer to a complete beta of our Unibright framework. Telling from our Roadmap this beta should be available until the end of the year, and we are well in time.

Working on a scaled agile development structure, we do not only use scrum as a framework for each team, but also implemented a ”scrum of scrums” on top of the 4 teams.

How does this look like?

• The “top scrum” defines an overall goal, for example our current goal to provide a public demo of our framework, showing all framework components working together
• Stories like user management (e.g. login once to use all framework elements with this login), persistance (e.g. saving a workflow design for later use) and integration tests are also defined on top level
• The different teams break down the stories on top level into developer tasks in their domain (Workflow Designer, Lifecycle Manager, Connector and Explorer)
• During a scrum, we also establish common daily reports and direct cross-team communication channels, e.g. on Slack

By our approach we are able to work with distributed teams that do not block each other. The teams stay agile and have enough room to organise themselves while still moving in the same direction and heading for a common goal — the production readiness of the Unibright Framework!