Data is an integral part of engineering at Student Beans. We rely on data to know we're building the right products, to find flaws in our systems, and to measure our company's performance.
In the same way, engineering managers use data to understand the teams' performance and to debug problems with their processes. We use DORA metrics, among others, to do this.
Below, we've given you 20 working days of data that should give you insight into the department's performance. You have three tables to work with: team membership, productivity scores, and merge requests.
Use this data and the python pandas library in a Google Colab to create a report on the following metrics:
Lead Time to Change - Average number of hours to complete one change (time between creation and deployment)
Deployment Frequency - Average number of deploys per day
Change Failure Rate - Percentage of changes that fail
Productivity Score - Average rating out of 5 for the question “How productive do you feel?”, asked each week
The data provides insight into three teams. You can assume that all teams consist of both backend and frontend engineers and that they all use agile methodologies.
The teams also have a leadership group made up of a Product Manager, a Delivery Manager, and a Technical Lead (Frank, Lucy, and Rob).
Make a note of any other assumptions you make about the engineers and their teams to share with us in the interview.
The teams' performance is your responsibility, and this data is your insight into their successes and challenges. Make a copy of this Colab, or create a new one to build a short report that interprets the data and explains any insights you've gained from it. The key things we would like to hear about are:
- Any successes the teams are having
- Metrics that indicate a problem
- Possible causes for those unsatisfactory metrics
- Actions you would take to help the engineers move those metrics in the right direction
In the interview, you will share your screen to show your report and take us through your findings. The report itself does not have to long or wordy (bullet ponts are fine), just enough for you to explain your workings and findings.
To start, make a copy of this Colab, run the code in the snippet to import your data, and build your report from there.