Share This Post

Nasscom Community

Cognitive Agents in Hiring

“You mentioned in your resume that your highest qualification is a masters degree in computer science – Is that accurate” Yes “Are you open to working in rotational Shifts” What’s that “Rotational Shifts means the employee has to work in a particular shift for a certain period and then they may be shifted to another shift. It means there is no fixed shift that the employee will be assigned” “Does that answer your query?” Yup “Moving on now… Are you willing to work in rotational Shifts?” Yeah “thats great to hear” “Are you willing to relocate” Yes…. Seems like a normal conversation doesn’t it? Between a recruiter and a candidate? And it is. Only thing is that the recruiter is not human. It’s an NLPBot! NLPBots are multifunctional cognitive agents built on the Cognitive process automation platform where they are trained on different data sets to perform various tasks on behalf of humans. Today we will see how they are impacting the Hiring functions. What are the main challenges faced by recruiters? I am not an HR expert but from our own experience I can say that it is critical to find the right person for the job at hand. In recruitment lingo that translates to the perfect match between the Job description or Key Responsibility areas and the candidate. This sounds simple enough but the key is in the details. These are critical across various stages in the typical Hiring Process. Job description: One of the most important parts in hiring and I am told also the most overlooked is the Job description. A good JD ensures that the hires are better. This is the first point in the hiring processes and if this contains any bias the whole process will be effected. In AI based Hiring processes automation it is critical that the systems prompts and collects all the data and points out any biases within the JD itself so that the process is optimised to convert the right candidate to an employee. Sorting – To find the right candidate the only window to the capabilities of the candidates up front are their resumes and there are enough data points to prove that almost every resume is not precise and in most cases highly exaggerated. This means that the AI has to identify whats true and fake when it parses a resume. There are key aspects that a trained recruiter notices within the resume that helps them validate the resume data and the AI is trained with those and other aspects to pull out those resumes from the pool. Now, the AI still has to match the JD to each resume and assign a score of how close a mayche they are and assign a score. Based on this core the resumes are ranked in order of relevance to the JD. Here the accuracy will definitely not be 100% (as in any AI system, that’s always learning). However the loss of accuracy can be counter weighted by spreading the net a bit wider and reaching out to a wider audience. Reaching out: Once the NLPBot has sorted the resumes it extracts the contact information from the resume itself and sends an SMS and Email to the candidates, this typically contains a link where the candidate can be screened by the Bot. Screening:  Here the candidate clicks on the link they received from the NLPBot and a chat window opens up. A typical conversation that ensues is shown above and in similar fashion the bot conducts these interviews with all the shortlisted candidates at their convenience and ensures that they are available, and willing to take up the job. Even other typical data can be collected and also shared with the candidate. Based on that conversation if both the enterprise and the candidate would like to progress the NLPBot displays the calendar of the interviewer to the candidate and they can schedule the human interview. Even the Human interview is full of data of the candidate and all the interactions the candidate had with the AI. This makes much more meaningful conversations possible during the interview. The AI can also learn from any other inputs (i.e. tech evaluations, skill mapping, etc) and make the human interview experience more productive. Thus we see how the an AI based Hiring process Automation that begins from when the business has a requirement and prepares a job description all the way to the human interview is completely automated due to these multifunctional cognitive agents (NLPbots). The efficiencies in Ai based automation is not just that we are able to hire faster and more efficiently but also ensuring that the “right candidate” doesnt slip through the cracks in the system.

The post Cognitive Agents in Hiring appeared first on NASSCOM Community |The Official Community of Indian IT Industry.

Share This Post