Defence Science and Technology Laboratory
MoD's Dstl (Defence Science and Technology Laboratory) wanted to support analysts producing high quality intelligence output, helping to find useful online tools in their various environments, including internally and online. They also wanted to start capturing usage data to make better decisions on iterating and supporting their internal applications and tools.
A thorough understanding of the users and their user needs based on insights from user research.
An initial backlog based on those user needs and the biggest areas of uncertainty.
A thorough review of an existing DSTL prototype to understand the quality/technical debt to decide on reusing the existing code.
Frequent delivery of working, high quality, tested Python code using Python, Django, Jenkins, Git, GitHub, Docker, and more.
Fortnightly progress reviews in show and tells with the product owner, and helped ensure decisions were made based on regular testing with users. This ensured we always delivered the most valuable work for the client, week by week.
Operational support to Dstl in deploying the system to their various environments (some of them not Internet connected).
Consultation on wider technical direction including proposed approaches to proxying, logging and authentication.
Regular (private) blogging for client and stakeholders about milestones, checkpoints, decisions made during the project.
A description of not just the system but our team principles and ways of working, to assist new starters as well as to ensure easier handover to future teams.
DSTL open sourced the resulting project, where the code can be seen at:
https://github.com/dstl/lighthouse (the app)
https://github.com/dstl/lighthouse-builder (code to automatically provision and deploy)