Easy start in IT: what is RPA and how to master the technology from scratch
Hello! I have been working at Intetics PM for 9 years. For the first 8 years, I was engaged in a project in the field of geographic information services (GIS), where we created coverage with high-quality geodata for a very large customer almost from scratch.
A couple of years ago, when the need arose to improve the efficiency of processes on a project, my boss threw up an article about the so-called robot revolution. The author talked in all the details about which professions will die out over time and why. Robotic Process Automation (RPA) technology was mentioned somewhere between the lines. Having become interested, I began a deeper study of this line of business. I really wanted to cheer up neurons, get out of the comfort zone and try something new, especially since AI and ML are actively used in GIS services.
In fact, automation can be done both in GIS and in procurement or financial sector. My transition to this direction began just with attempts to roboticize the current front-office tasks of a GIS project and internal back-office processes of the company.
We trained on cats for a long time, on simple routine tasks. For example, together with the recruiting department of the company, we have long and persistently created a bot, which was taught to pull out the resume of candidates with a certain technology stack from the database, form these resumes according to the company's template and give the recruiter a ready-made resume for work.
All of our experimental bots have performed well within the company and have really reduced the number of manual operations. This is how we became the RPA services. At the start, there were two of us such enthusiasts, during the year 5 more people joined us (including those from the GIS team). We have already launched several client projects and the team is now actively growing. The technology is quite new, few people know about it, so we also have to carry out popularization work :)
What is RPA
Robotic process automation is the use of technology to automate business processes. By building algorithms based on a special platform, the developer gives the robot clear instructions and tunes it to perform the necessary tasks. And with the addition of machine learning functionality, the instructions may become less clear, and the robot has a certain freedom of action.
Automation can include data processing, interactions between different digital systems, and more.
One example of a real-life project that interns at the company learned to write robots is a music event site. Previously, this very collection was created by people who googled events around the world, translated the description and posted on the site. The robot, in fact, freed the company's employees from routine searching and posting. Now a robot with integrated artificial intelligence looks for such events and translates their descriptions into the language of the site.
Another cool example of using a bot came up with one of the largest car companies. When buying a car, the bot will contact the buyer in any convenient way (SMS, messenger, email), list a list of its services, and also provide any assistance in real-time mode. For example, you don't know what kind of light is on on the dashboard? The bot will help you figure it out, give you advice on when to change the oil or refuel, and may even book you for repairs or consultation.
Some people think that RPA is synonymous with AI (artificial intelligence) and ML (machine learning), but this is not the case. In robotization, the robot cannot deviate from the set rules and strictly follows them - this is the essence of RPA. The essence of AI and ML technologies is to train a machine to make decisions autonomously, deviating from established instructions.
Nevertheless, robotization can include elements of AI and ML - and these are the "hottest" technologies in the modern IT world.
Why is this technology trending
All this sounds pretty impressive, but you need to add a little specifics to understand why the world's largest companies have begun to actively implement the RPA approach into their production processes.
Robotization eliminates the possibility of errors when performing tasks. Since the program is given certain conditions, it cannot fail to fulfill them - unlike a person who may not notice a small nuance that will further affect the entire process.
Finally, robotization also helps reduce costs for the company. Having once invested in the development of a high-quality robot, in the future the company will not need to pay it severance pay, wages and other items of expenses, which are mandatory in the social package of any employee.
And also, in comparison with the development of special software, RPA provides fast implementation and, accordingly, a quick return on investment, but at the same time it does not require restructuring of existing processes and IT systems of the company.
RPA refers to a low-code approach to development: low-code, in turn, is the use of ready-made modules to create specific solutions. This allows you to develop software with minimal manual typing and automates repetitive tasks.
This approach brings new and promising professionals into the industry who do not yet have sufficient programming experience, and for the pros it is a great chance to switch to more creative tasks. This means that a person with basic programming skills (or without them at all) may well master RPA and not become unemployed junior testers and business analysts.
How to become an RPA developer from scratch
When hiring for a job as an RPA engineer, they pay attention primarily to algorithmic / process thinking.
But there is a certain set of skills that are still needed if you want to try yourself in RPA. It:
- Knowledge of Microsoft Technology technologies (VB .NET, Windows, SQL Server, Web Services, MS Office) at a basic level.
- Understanding the rules and principles of analysis, design, development, implementation and support of code in different contexts - but a basic level (for example, watching video tutorials) will be enough.
Typically, for engineering students or graduates of IT courses, these requirements will not create any problems. Other occupations are likely to have minimal initial training.
And if you want to delve into technologies such as machine learning and artificial intelligence in the future, as well as more "advanced" RPA, then, firstly, you need to study the most famous RPA platforms and tools: Automation Anywhere, UiPath, Blue Prism, Softomotive, Kryon Leo, WorkFusion, etc. Second, learn how to write code. Basically it is recommended to learn Java, you can also learn Python, .NET, C #. It is not necessary to be able to code from the very start, but if you want to delve into RPA, you will have to learn one of the programming languages in order to expand the existing functionality of the RPA platform if it is no longer enough.