There are a few ways to build a Python RPA automation, but it is important to consider quality and best practices from the beginning of this build to minimize potential problems. Let’s learn about some of these best practices below.
1. Have an open communication with the people involved
It is important to make it clear that automation was a perceived opportunity to improve the execution of a manual process, and that the person who was previously responsible for this same process will now have more time available for new tasks that demand their knowledge, intelligence, and decision making. It’s necessary to talk about the culture of automation so your teams understand that they are not losing place to a robot. Read more about it here.
2. Get to know the process
And when we say that knowing the process is important, we mean not only what happens within the execution of that process, but also what happens before and also after.
Take the time to understand:
- The step by step process when everything happens successfully;
- The conditions for everything to happen successfully;
- The mistakes that can happen during the execution of a process and have them mapped out in your planning;
- The input data and the output data;
- What are the rules of these datas;
- Which people participate from one end to the other;
- What systems are involved in this process;
- What access or environment permissions these people and systems need to have;
- How long this process takes to happen before automation, considering whether each part occurs in sequence, or on different days or at different times;
- Among others.
3. Document the process
Besides understanding how a process works before automating it, it is extremely necessary to document everything that has been mapped. This also helps in the planning and maintenance of the built automation.
4. Communicate changes in the process
The built robot will obey what it was designed to do. If, by chance, there is a change in some type of data used in the process, or one of the systems used changes some rule or field, this will make the robot to stop working. Not because it was built the wrong way, but because it was not updated. So it is important that this communication happens clearly and as soon as possible.
5. Update the documentation
And the previous point also leads us to bring in this practice of keeping the documentation up to date. This way, whoever needs to understand or do some maintenance on the automations created for those processes will have the up-to-date and correct information to consider in their actions.
6. Have a well-defined flow for your automation development
Besides developing an automation, it is necessary to have a flow for this development to occur in the best possible way, with well-defined steps, controls, and follow-ups, trying to minimize possible post-production problems.
Here are some tips:
- Make sure that the person or team responsible for creating the robot has understood how the process works;
- Consider quality during and after development by performing some tests, like functional as well as security and integration testing;
- Have separate environments:
- a development environment, a testing environment, also known as the homologation environment, and finally the production environment;
- Have a procedure for reviewing what has been built, also called Code Review, a practice widely applied by many developers;
- Among others.
7. Avoid Shadow IT and ensure automation governance
Automation projects have become increasingly robust, with concerns similar to a software project, such as application and data security, governance, scalability, and observability, among others. It is recommended that you have support from the company’s IT department so that they have the knowledge of what is being built, they can bring in the points mentioned, such as security, governance, and others, and help with maintenance, if necessary.
8. Orchestrate your automation
Keeping track of how the robots work, especially when there are several, can help you to monitor these automations. Not just ensuring that they have or have not been executed. But understanding possible errors, at what time each robot should be executed, in what sequence, etc. This makes day-to-day life easier, allows you to act where possible problems might occur, and further improves processes that have already been automated.
Orchestration can involve both the business area and the IT team (or CoE). And some features such as automation scheduling, alert triggers for users (process started, system down, error in request), management of logs from the processes created, integrations with communication systems (Microsoft Teams, Slack, Whatsapp, and E-mail) can guarantee a scalable, centralized, and efficient operation in your company.
In our orchestration tool, for example, you can track these alerts in the way we described above. In the image below, we have a demonstration of the screen with several options available in the menu located on the left side, and one of them is identified as “Alerts”. Clicking on this option opens the central screen on the image with one alert per line.
How has it been for you or for your team? Which best practices have you applied in your daily life? Share with us! And you can learn more about our solutions and how to better orchestrate your automations, even if they are already running in production, by clicking here.