Data science is a misnomer because data is at foundation of any science. However many teams striving for agility lack scientific method and are prone to witchcraft and charlatanism, trying to replicate anecdotes of successful agile transformations. I'm sure as an industry we can do better. We have many sources of data describing our work. If used wisely, the data can provide insights into the trends and health of the change process. Data can also give grounded evidence of areas worth focusing improvement on.

In this session we will review what data is available to the teams today inside the well-known systems of record (issue tracking, source control, CI etc.) and what data is worth starting to collect. We will see how to combine multiple data sources together and present the insights in a meaningful way, that tells a story. We will learn how to look for patterns in data and how to tell if those patterns are real.

For example, a source control system "knows" what files change most frequently and who changed which files (and even lines of code). Combined with the issue tracker data, it can tell you the ratio of "lines spent" between bugs and features or if your "effort per feature" is trending up or down. Now you can measure your technical debt per feature area or system module!

Come and learn what story can your team's data tell you and where to focus your continuous improvement effort.

 
1 favorite thumb_down thumb_up 0 comments visibility_off  Remove from Watchlist visibility  Add to Watchlist
 

Outline/structure of the Session

  1. Present various data sources available to the teams
    1. Share the most common sources
    2. Briefly survey participants for more ideas
    3. Share some less-known potential data sources
  2. Share techniques how the data can be extracted out of the well-known systems and which formats are preferrable
  3. Demonstrate (briefly) Data Science 101 techniques of loading, manipulating and presenting data (with easy visual tools, like Tableau)
  4. Show common agile tools with analytical features out of the box (Jira, Jenkins etc.)
  5. Show how to correlate data and what to look out for to avoid misleading correlations
  6. Give outlook of research in the area and where to learn more

Learning Outcome

  • Learn basics of data science
  • Get inspired of the potential your data contains
  • Learn practical techniques how to tap that potential tomorrow in your agile work
  • Understand how to apply scientific method for change management and agile coaching

Target Audience

Teams, ScrumMasters, coaches

schedule Submitted 1 year ago

Comments Subscribe to Comments

comment Comment on this Proposal