Deployment time reduced from 1 year to 2 weeks at public travel organization

Devoteam performs numerous DevOps implementations for both businesses and (semi) governments. Every organization has its own problems that need to be solved. But one thing is certain: the goal is always to optimize the IT processes of an organization through automation. How do we proceed?

This public transport organization provides a service that must seamlessly meet the needs of its target group. With multiple means of transport, many employees, millions of travelers, and multiple routes totaling thousands of kilometers, this transport organization has an important function in the daily life of almost everyone. In this playing field is data crucial. Optimizing operational excellence plays a major role in continuously improving and realizing the travel vision.

The challenge: from a lot of different data solutions to one data foundation

The longer it takes to transform data into insight and information, the more value it loses. In addition, a large amount of databases, warehouses and other tools ensure that there is often not one truth. There are multiple truths, depending on which data warehouse the insight comes from.

The demand for a robust data platform meant that a gigantic street had to be set up to automatically retrieve data from various data sources of the provided service. This includes information from travelers, information from different means of transport, routes, travel times, delays and endless other sources. An example of a report that must come from this street is an annual report to a Ministry stating, for example, how many different means of transportation arrive on time each year at stations. No release to production could be done for a whole year. This was no longer sufficient and of course had to be done more often and faster.

The solution: DataOps – Apply DevOps pipelines and automation to the Data Platform

During the development of this new robust data platform, expertise was needed to automate the deployment street with a lot of Azure software and pipelines had to be built. In addition to the technology, it was also the intention to include the data team in the DevOps principles and thus a DevOps way of working. Projects and activities had to be cut into smaller pieces and dependencies on other teams had to be handled differently.

Cutting up parts of data also had a great advantage for dependencies in the data domain. Instead of one big box of data, everyone gets their own piece. As a result, dependencies between teams became smaller, so that teams can also work independently. The pipelines and the entire street must then reconnect to this.

Technique and process

On the technical side, Devoteam contributed to the development and implementation of the street, the realization of pipelines, the exploration of how processes can be automated and ultimately the actual automation of these processes.

People and culture

Devoteam also played a role on the culture side of DevOps to ensure that the data team was familiar with certain DevOps techniques and the Agile way of working.

The result: From deploying once a year to every two weeks

In the beginning, deployments took place once a year. Currently, a deployment is done every two weeks. The goal is to be able to do this on a daily basis. This still requires a number of technical and procedural changes to be made. For example, even more overview and structure must be created in the process and elements such as checking the database must be accelerated.

What is the result of these deployments? Power BI reports for this organization itself, but also for Ministries, etc. This enables the organization to act faster with data-driven insight.