Closure Report

Closure Report

Project Summary

This project managed developments to be made to the Peace Agreements web application and database PA-X (https://www.pax.peaceagreements.org/) for 22/23.  The application has been developed over the last 10 years by Development Services to support and publish research carried out by the Political Settlement Research Programme, (now PeaceRep), and the School of Law.

Prior to HSS042, project activity had been managed directly by Development Services in close collaboration with the PeaceRep team. In order to support an ambitious and expanding roadmap of development work, and to provide a consistent and reliable level of service to the research team, we have moved to deliver this and future work within our standard project methodology and governance processes.

Scope

Objectives and Deliverables Achieved

The project was initially up against a tight deadline to deliver improvements by (research grant) end of financial year March 2023, which led to the project being given a higher priority. The initial backlog of Jira items was organised and prioritised and the must-haves were delivered by the deadline. The project was then extended to the end of July 2023 and a contractor recruited to provide the additional development resource required to deliver planned improvements ahead of this (Applications directorate) financial end of year deadline.

So far, the development work had been executed in ad-hoc "sprints" of activity, and this project established more formal agile procedure around them by introducing regular sprint review and backlog prioritisation meetings with the project team.

To support onboarding the project team to the methodology, a workshop was held to define the goals, provide context, and initially prioritise and organise the Jira tickets. Following the workshop the teams were better able to understand how the individual developments fitted together in the best order to achieve the end goal, and work was parcelled into "bundles" that made sense to everyone rather than "sprints", as recorded in the change log (Change in execution method). The workshop also helped to build relations between PeaceRep and Applications Directorate teams, many of whom had never met.

The primary objective of the first bundle of work was to migrate the Peace Agreements website from the legacy PHP platform onto the Django platform.

The budget was increased by 45 days which allowed development work to continue through July 2023. Around this time, following the delivery of bundle #1, the project team had to make an important technical and strategic decision on whether they should move ahead with an automated text matching approach when migrating agreements from the legacy application into the new Django platform.

Around the same time, in response to prioritisation decisions, the team shifted focus to develop and deliver a new Tracker application. Tracker developments were completed, and this was launched along with the new migrated Peace Agreements site in September 2023. 

Delivery of the tracker website including a public launch allowed the team to meet a funder requirement to bring this website online by July 2023. The project then extended its Delivery and Closure milestones to the end of October so that more improvements could be delivered while the contractor was available. The development team went on to develop, deploy and maintain the 'Ceasefires in a time of Covid-19' website. 

As previously mentioned, a major deliverable for the project was the migration of the legacy PHP Peace Agreements website, and replacement with an application on the more modern Django infrastructure, with enhanced functionality. This was launched to the public in March and April 2024 with the following features completed :

• Design and deployment of a new interface for tagging agreements
• Migration of the data from the old system to the new architecture
• Replication and enhancement of the search pages from the agreements database (still undergoing testing and awaiting feedback from PeaceRep ahead of final improvements)

The project ran as a development led project between October 2023 and February 2024 as the Project Manager left and there was a gap before the project could be picked up by another Project Manager. After the new Project Manager took over in February 2024, the decision was made to keep the project open until the end of the year (July 2024) to allow for the completion of the work on the replacement for the PHP site and to cover post-Live issues.

Benefits

Success Criteria

Analysis of Resource Usage -

Staff Usage Estimate: 100 days at end of planning (Feb 2003)

Staff Usage Actual: 445 days

Project Services - 48 days

Software Development - 380 days

Development Technology - 12 days

Apps Management - 4 days

Tech Management - 1 day

Explanation for Variance

Initial budget was 100 days to end of March 2023. The project was extended to end of July 2023 and the budget was increased to 125 days. The project was extended again to end of October and budget increased to 175 days. Contractor time was in addition to this.

The Contractor in place became a permanent member of the Software Development team in Autumn 2023.

Due to the Project Manager moving on, and a gap before another Project Manager could take over, the project ran on as a development-led project for a number of months.

The budget was updated in January 2024 following a new Project Manager being assigned.

The Developer left the University at the end of October 2024. The follow-on project was started up to coincide with this departure to close this long running project.

Outcome

What started as a short-term project to deliver something quickly became an ongoing support structure to help deliver improvements. Developments were delivered more quickly and more flexibly and the teams worked together more effectively. The project has been a success, delivering a consistent developer resource to achieve many important improvements and features, and maintaining flexibility around prioritisation of work.

Key Learning Points

Log link

Outstanding Issues

• The follow-on project will continue to address the improvements that are prioritised for 24/25 and will continue to use the same Jira backlog as this has proven to work well with the team
• The follow-on project will be planned with dates to run in tandem with the research grant financial year, to simplify resource planning and budget management.
• Handover to Production Management will carry on to the next project as they become more involved in releases to live.
• Decommissioning tasks related to the new website going live will be one of the objectives from the next project
• The Jira Backlog is used to record all improvements and bug fixes and is used to drive the 2-week sprint cycle
• The implementation of new semantic similarity technology, to reduce workload when tagging agreements, will continue into the follow-on project