She adds, “If a company has to hire 10 people, I have to go through 100 résumés, and a week goes in sorting out the profiles. Now, an AI system would do that in half a day. In terms of the career advisory kind tools, within our organisation also we have chat bots in terms of our specific competency areas to figure out what’s the latest training available. So, that happens in an instant. You don’t wait for the learning leader of the group to come.”
Teaching AI machines in terms of stickiness of hiring is still in process, Mohapatra says, but one needs the collective intelligence of the best recruiters to create an effective base model. “When it comes to assessing skills and plugging gaps, in terms of matching job profiles with roles, the process is on. In terms of tone, personality, etc. we have an Application Personality Interface called Personality Insights. So, if you feed in a person’s CV and Twitter, LinkedIn posts it will give you an insight into her personality, where you will know whether she will stick on.”
Neeraj Sanan, chief marketing officer of Spire Technologies
, which offers contextual (AI) products for HR functions, cites two inefficiencies in hiring — first, lack of jobs and candidates at the same time; secondly, the entire recruitment process in any organisation in India takes in excess of 125 days. “Technology should be capable of reading job descriptions and résumés, and matching A to B. The concept of reading a résumé like a human... we call it unstructured data comprehension or textual data comprehension.” He adds that the machine should be able to compare two individuals with different language proficiencies. Also, the AI machine must be able to analyse various kinds of documents in one go.
“When a software is intelligent enough to give results contextually to each individual, this is called contextual analytics. We do contextual analysis on text data. AI is the root to that. The power is that you get output contextualised to you.” Such a software can self-correct its prediction capability, which is what machine learning is. Spire combines two skills — comprehending unstructured data and contextual analytics.
Prasad Rajappan, managing director, ZingHR, a Microsoft partner that provides recruitment solutions for organisations, says innovative tools like AI is being used in many HR processes. “While tech companies
seem to be the obvious adopters, we are seeing non-technology companies, especially BFSI and retail, adopting them more proactively,” he adds. While established tools like Microsoft’s Azure Machine Learning help build predictive model for intelligent hiring, Rajappan feels AI revolutionises the recruitment process by applying complex mathematical models on sound HR domain expertise and established HR models to predict the success of a candidate. “AI models help eliminate interviewer bias while helping achieve process efficiencies by over 80 per cent,” he says, while recommending that it is imperative to have a strong mix of professionals who are sound in the HR domain and understand mathematical models related to business applications.
Sudheendra Chilappagari, co-founder of Belong, a Bengaluru-based start-up that uses big data and analytics to aid recruitment across industries, says new-age AI-powered software is going to automate all manual tasks and free up time for recruiters to concentrate on engaging candidates. “Companies
are also realising that the best candidates today are no longer applying for jobs, and that’s escalating key metrics like time to fill, cost per hire, quality of hire etc. This is why many talent acquisition teams are actively deploying analytics-driven recruiting to target and engage top candidates wherever they are.” He points out that high-growth companies
Belong works with actively look for “an entrepreneurial DNA in their candidates”.
The start-up’s internal research shows it is able to increase recruiter productivity by up to 50 per cent through automation and machine learning. “This is largely because recruiters spend too much time either searching or interviewing candidates who are not a fit for hiring leaders. Our Adaptive Search, however, looks at a company’s sourcing and candidate engagement patterns, and their current and past employee data to identify candidates who are most similar to their successful candidates or high-performing employees. When this learning is fed back into search, we are able to reduce the time recruiters spend on search by as much as 50 per cent.”
In terms of application, Sanan of Spire points that such a technology is vertical-agnostic and domain-agnostic, so it can be adopted industry across sectors.
“We start with an Internet-based intelligence. If the functional manager is not following our prediction, the software keeps correcting itself so that in about a few days from prediction our software keeps mapping as to what exactly the manager wants. It’s not a hard-coded algorithm which when once written cannot be undone. Algorithm is self-mapping actual behaviour, and that’s the power of artificial intelligence. Only if the algorithms are hardcoded, it may work in industry A and not in B.”
The new-age technologies use machine learning with a capability of being transient because for the same job over a period of time one is demanded newer skills. “A true-blue AI product using machine learning like the way we’ve built will be able be adaptable enough to take care of changes over a period of time and be able to give differentiated results the moment the technology or the industry changes,” he adds.
Making the most of AI
* AI solutions offer the best profile matching
* They help eliminate interviewer bias while helping achieve process efficiencies
* It is imperative to have a mix of professionals who are sound in HR and understand mathematical models related to business applications
* Research shows that AI can increase recruiter productivity by up to 50 per cent through automation and machine learning