Have you ever thought about ways of reducing repetitive, monotonous tasks? Maybe you would like to try to automate your own tasks? I will show you what technology we use at Allegro, what processes we have automated, and how to do it on your own.

What is RPA and how do we use it at Allegro?

“Robotic process automation (RPA) is a software technology that makes it easy to build, deploy, and manage software robots that emulate humans actions interacting with digital systems and software. Just like people, software robots can do things like understand what’s on a screen, complete the right keystrokes, navigate systems, identify and extract data, and perform a wide range of defined actions.”
Source: UiPath Robotic Process Automation

At Allegro, our Process Automation Team primarily relies on UiPath as our key RPA tool. Processes suitable for automation through RPA are standardized, repetitive, manual, with high volume, stable and has data in digital form. If possible, we try to combine UiPath with different integrations such as scripts, databases, chatbots.

UiPath provides the ability to automate all kinds of applications (web, desktop, java, etc.). Automations can be created through the user interface of an application, meaning that the created robot imitates an employee’s clicks. Furthermore, when an application’s API is available, it is easy to integrate UiPath with API, and in that case, the robot’s steps are performed in the backend-side. UiPath also allows the use of OCR (optical character recognition) and machine learning modules.

Thanks to the various roles within Process Automation Team, such as analysts and developers, we are able to approach processes holistically. When we receive an idea for automatization, we first perform an assessment to establish if the process is suitable for robotization, and we calculate the Return On Investment (ROI) and the potential of time savings from automation in terms of Full-Time Equivalent (FTE). Once the assessment is done and costs of investment return are approved, we analyze and optimize the process. As a result, the analyst prepares a Process Definition Document (PDD) which serves as an instruction/description of the process. In the next phase, based on PDD, a developer takes over the process and designs a solution. After that, the development part begins while the robot is created. Last but not least, there is the testing phase, where we check the results of the robotization together with the analyst and the business process owner. If the solution is successful and performs as intended, we run the robot in production. Then we enter the hypercare period, during which we monitor and make necessary adjustments in tandem with the business process owner. After about two weeks of this phase, if both sides are satisfied with the results, we “go live”.

Robotic Process Automation Workflow

What kind of processes do we automate?

Jira automations

In our organization we have a lot of processes based on Jira “tickets”. Many employees had to manage and operate Jira’s queues. Our team implemented several robots to relieve administrators from repetitive tasks. Jira has an API available which we used in combination with the UiPath platform. For example, when an employee is leaving Allegro, several Jira tasks are automatically created to retract authorization in various systems. Previously, these tasks were performed manually by administrators. Now, the process is fully automated. The robot manages tickets via API and checks accounts in systems by GUI.

SAP (Enterprise Resource Planning system) processes

All repetitive, rule-based processes in SAP can be automated. For instance, let’s consider the processes in the Finance team. They handle massive amounts of invoices. For some suppliers, with the largest quantity of invoices and unchangeable invoice layout, we were able to automate the accounting process in SAP. The robot accesses an appropriate transaction in SAP and lists invoices. Based on business rules, the bot selects a specific invoice, opens it, and reads selected values. Then, it compares these values with business conditions which were established in the Process Definition Document. Depending on the situation, the robot fills or corrects adequate fields and either accepts or rejects the invoice.

Automation by API

One of the excellent examples of API automations is the process of changing product categories on the Allegro platform. Allegro hosts a vast number of products. Initially, not all of them are assigned to the proper category. We were able to create a robot that uses Allegro’s REST API to move these products to the target category. Before automation, this task was time-consuming and monotonous. Recently, the robot completed a huge task, moving almost 3 million products in two days!

Processes across multiple applications and integrations

It is possible to combine tasks from different applications into one robot. This approach allows us to automate more complex processes. The most interesting ones include:

  • The process of user’s data change in Allegro platform is carried out at the request of users. To perform all actions, the bot uses Salesforce and Internal Admin tools. The robot downloads a report with requests, then checks the pre-set business rules. Based on the results, the bot changes user’s data or rejects the request. The robot works 24 hours a day, handling 80% of applications. The number of tasks it performs can be compared to the work of four full-time employees.
  • The anti-fraud process. The robot verifies hundreds of thousands of messages and blocks suspected accounts. Using suspicious messages from the Spoof application (Message Center), the robot determines if a message is spam or not. To make a proper decision, it checks various business conditions to decide if an account should be blocked. After making the decision, the bot blocks the account’s message sending capabilities.
  • The process for the HR team where the robot works on two applications. The robot interacts with the interface of an application and also uses its API. In the recruitment process, specialists from different fields participate and help recruiters to find the best candidates. These specialists are known as the Hiring Squad. A significant number of people are involved in this process, and the robot is responsible for keeping the Hiring Squad updated. Based on a report with job offers the robot checks if a candidate has active status for specific skills required for the interview process. If the status is active, the bot selects a particular job offer from the platform and assigns the interviewer from Hiring Squad.

“Human in the Loop”

Those processes which are rule-based, repetitive, but require human decision-making or the robot does not know what to do on the basis of the collected data, are referred to as “human in the loop”. A great example could be the process of damage complaints regarding packages that we have automated. The robot gathers a report from Salesforce and filters all jobs referring to damaged packages. Then, in the Internal Admin tool, the bot checks and collects various pieces of information based on specific business rules. If necessary, it also checks the status of packages on carrier websites. Finally, the robot creates a form with all the gathered details, information, and attachments. This form is sent to a human for verification. With all this collected information, an employee can quickly decide whether the complaint should be accepted. Then, the decision is sent back to the robot, which is able to close the case. It sends appropriate emails and records the results in the system.

Robotics workshops for employees

Additionally, twice a year, we organize an educational program for employees called “Allegro Robot School”. Employees get a chance to learn how to build basic robots in UiPath and build one to support their daily tasks. To sign up for the program there is no need to have coding experience. It is enough if an employee can think analytically and has motivation to learn new things. The program is intensive, consisting of five days of workshops. After the workshops, there is a three-weeks period where, with our support, employees choose their own processes and build robots. For each edition we have around ten participants. The robots created during one edition account for about 3 FTEs! We have many examples of graduate employees who created more robots to support their daily work in a team. Moreover, we created a Slack community for graduates to stay in touch, share knowledge, and support the development of new robots.


Today, the number of processes and different applications used in companies is enormous. Moreover, it can sometimes be challenging to integrate one application with another, and employees are burdened with many manual, repetitive tasks. It is important to know that there is a solution to automate these processes. The RPA technology can quickly help with that, freeing up employees for more creative tasks. The above examples visualize the possibilities of using UiPath. The most important thing to remember is that the process to automate has to be manual, rule-based, repetitive, with data in digital form. What is more, it is possible to learn for those who don’t have coding experience. Thanks to that, the automation of processes can be expanded across the company beyond the Process Automation Team.