In the 'traditional' approach, there is a strict separation between different aspects of a project: there are separate teams for development, testing, system administration and more. These teams interact with each other to deliver infrastructure for deployment, deliver a release from development to testing etc, but each team largely operates on its own. Within these teams, a lot of actions (e.g. infrastructure installation and configuration) are performed manually. As humans we make mistakes, which makes these manual actions risky.
DevOps is an evolution of software development, and is a combination of software development (Dev) with system operations (Ops). DevOps combines the verious teams and allows infrastructure installation and configuration to be managed as code, which includes the deployment and testing of instrastructure as software releases. By managing system management as software, new installations or changes to existing installations can be performed much faster and at higher quality than through manual actions.
Next to automating systems and infrastructure management, the same process can be applied to deploying, testing and monitoring of data pipelines, analytical process and data science models (DataOps). An advanced DevOps and DataOps strategy allows to keep detailed control over processes and at the same time quickly respond to new or changing business requirements.
More than the purely technical impact, DevOps means a change in mentality. Merging development and system operations results in a much tighter integration and cooperation. It unites the agile way of working, continuous integration and delivery, automation and more to make the entire project life cycle more robust and more efficient. DevOps allows faster innovation and adds more value to internal and external customers.