Our Services

Solution

The project consisted of the following stages: investigation, PoC development and testing, feedback gathering, and PoC improvement.

Investigation

First, we needed to determine a business process that will benefit from automation. We established the following set of criteria:

• The process should be predictable, repeatable, and rule-based.
• There shouldn’t be too many exceptions.
• It should be manual and intrinsically prone to human errors.
• The process should be universal and industry-agnostic.

After interviewing heads of Itransition’s departments, we decided to automate the process of adding new candidates to our human resource management system (HRM). This is a standard process for many industries, which is predictable, repeatable, prone to human errors, and requires manual input. 

The process goes as follows:

• HR assistants search for candidates on LinkedIn and similar sources.
• They send an email to a talent acquisition manager to create a new candidate in our Zoho-based HRM system.
• The talent acquisition manager receives an email with a short description of the candidates to be used in the HRM system (name, surname, skills, status, etc.) and attachments with CVs.

This process met all of our automation criteria. Itransition’s HR managers add 20-50 new candidates to the Zoho database every day. On average, it takes about 2-3 minutes to process candidate data. The process doesn’t require subjective judgment and has a clear step-by-step flow. There are only a few application exceptions such as login failure, application not responding, and application timeout.

To identify requirements for automation, our RPA team conducted a set of interviews with Itransition’s HR and talent acquisition managers. Together we discussed process steps, exceptions, and what main applications (Outlook, zoho.com, Excel, etc.) are involved in the process. We also agreed on the expected results of process automation — to minimize human interactions and reduce candidate application processing time.

Based on this data, our business analysts created a process design document. The document describes the sequence of steps performed as part of the selected business process, the conditions and rules of the process before automation, and how the process should work after automation. Our specialists also created a detailed process map, which outlined the scope for RPA.

PoC development & testing

Leveraging the process design document, our RPA developers compiled a solution design document, which outlined all actions a robot should perform. Our team also defined operational arrangements (how often the bot should run, what time the bot should start, etc.) and exceptions handling (for example, if there’s no specific information in the candidate’s CV, the bot sends a ‘CV does not contain the necessary information’ message).

Being a UiPath Silver partner and having certified specialists on board, we opted for UiPath Studio for RPA implementation. Using the solution design document, our RPA developers added the required sequence of activities in the UiPath Studio.

Thanks to the RPA, HR specialists now only need to email the list of candidates with basic information along with the CVs attached to the bot. The bot does the rest of the job by parsing the candidate’s CV data, logging in to the Zoho database, checking whether the candidate is present in the database, as well as creating or updating candidate records.

 At the end of the process, the bot sends a comprehensive Excel report to an HR manager. The report outlines all the data that the bot retrieved and filled in, which helps with monitoring the bot’s activity and resolving issues. To mitigate false data entry, we created a set of databases with the names of the most common universities and faculties for the bot to check against.

To verify that the automation solution was developed according to the defined business rules, our team performed testing. We created 100+ test cases using different data sources (LinkedIn, zoho.com) and CV structures. First, we tested the solution in a sandbox, documenting defects and analyzing results. Then, together with HR managers, we performed user acceptance testing (UAT) in the production environment, ensuring that the delivered solution met their expectations.

Using the results of the UAT, our team created the list of new features and improvements, which formed the scope of work for the next project stage. Feature releases followed by stabilization releases ensured the bot’s flawless performance.

Feedback gathering and PoC improvement

When the solution’s beta version was ready, we prepared a demo for our HR managers to demonstrate the capabilities of the bot. During the demo, we collected feedback on the bot’s performance, discussed the ways we can scale the solution’s capabilities (for example, adding new sources the bot should use to retrieve information), and improved the solution in two ways:

• We minimized the time spent on manual data review by including all the critical information in the bot report email’s body.
• We implemented bulk candidate records processing via the Zoho import feature, minimizing the number of the bot’s actions and increasing its reliability. 

To ensure that HR managers can use the bot to its full advantage, we provided training sessions and created a user guide that explains the bot’s capabilities in detail.