Blog

“Automate everything” is more than strategy – it’s a mindset

Marko Tolvanen Delivery Lead, Solita

Published 30 Sep 2024

Reading time 3 min

Is the mantra “automate everything” a universal truth, a misconception, or simply consultant jargon? In my experience, it is a mix of all three. It’s true that automation has become an integral part of DataOps, aligning seamlessly with Lean principles by reducing waste and increasing efficiency. Yet, it is also somewhat misleading, because automation requires an upfront investment—time, effort, and sometimes new tools or licenses. Not every task is worth automating. 

Automation minimises manual work and improves data quality

Within data development, significant gains have been made by automating daily, repetitive tasks. Data engineers are automating processes such as CI/CD pipelines, Infrastructure as Code (IaC), ETL workflows, and leveraging orchestration tools. These automations minimise manual labour, reduce human error, and speed up complex processes—ultimately improving the efficiency of development and release cycles.

However, one of the most time-consuming aspects of data development remains data quality management. Most production issues are rooted in incorrect, missing, or delayed data, which requires considerable time and effort to investigate. Moreover, poor data quality erodes trust in the offered data solution. Fortunately, automated monitoring solutions help maintain data quality and allow for more proactive maintenance.

During the development phase, it is essential to identify specific data that can be monitored with business logic tests. For example, if a product’s price falls outside an expected range, an automated alert could trigger a notification to the ITSM tool, prompting the DataOps team to investigate. Similarly, we can set up automated checks for missing data, null values, or delayed data. By maintaining the mindset of “automate everything,” we can systematically reduce manual tasks and improve system reliability.

Start small with automation

If you are new to automation, start small. Pick a simple process for your first automation project. Define clear goals, determine how you will achieve them, and identify key metrics to measure success. Your target should be specific enough to assess whether automation truly made an impact.

A common rule of thumb is to keep automation as simple as possible. This is how you avoid the trap of overcomplicating things. While this advice is valuable for beginners, we’re increasingly seeing more complex processes being successfully automated, especially with the integration of AI. As tools become smarter, they’re able to manage more sophisticated workflows.

Automation benefits service management reporting

Automation is not only for data engineers. Service management can benefit significantly from automation as well. For instance, many service or project managers still manually create monthly reports. Modern ITSM tools allow you to automate report generation and schedule them to be sent automatically—say, on the first day of each month.

If someone argues that it only takes 30 minutes to create a report manually, remind them that over a year, that’s six hours of work. If it takes just three hours to automate the report, the return on investment is quick and substantial.

What’s next for automation?

The next wave of automation is already on the horizon. Machine learning solutions are emerging that can automatically fill in missing data fields, and soon AI will be able to resolve common production issues without human intervention. Generative AI (GenAI) is also being integrated into systems to automatically generate quality reports for delivered services.

Ultimately, “automate everything” is more than just a strategy—it’s a mindset. If you consistently seek opportunities to replace manual work with automation, you’ll gradually reap the benefits: more time for higher-value tasks and a more efficient workflow. 

  1. Business