MapRecruit, President and Co-founder: There Should Not Be Any Bias in Data
(US and Canada) Swamy Sriperumbudur, President and Co-founder, MapRecruit speaks with Mark Johnson, Regional VP, Fusion Alliance, about the role of HRtech in the transformation of organization and how automation can reduce biases in data.
The company was named MapRecruit to keep it simple. While the ‘recruit’ is all about the end-to-end talent acquisition platform using AI, the ‘map’ comes from mapping the core functions.
The first function is to get the correct job description. “That is where we have developed our knowledge graph and we have created so much deep-dive information on each of the different demands to make sure the right role is shared with the job seeker. We created something called a source AI, where you have all the different elements of the sourcing — who, what kind of a person, what would be the elements of parsing for JD, and all the attributes. We are coming down to the attribute level because we are going to the deep insights of each role,” Sriperumbudur says.
MapRecruit also has a matching and ranking function to find the best match for a particular job description from many suitable resumes. The resume parser connects with the JD parser and looks at this attribute level to match the required skills with the skills of a job seeker. The platform’s Engage AI kicks in after identifying qualified candidates, to engage with them properly.
“Once you engage, you need to pre-qualify candidates. And if you do that using our AI and ML, you put them in a bucket where hiring managers can do a complete qualifying interview. The entire workflow processes have been mapped in the recruitment process,” he adds.
Speaking on the condition of human-machine bias in ML and how MapRecruit is avoiding historical bias in the data sets, Sriperumbudur says, “There should not be any bias at all. We need signs or reasons and cooperation to drive out ignorance and bias. We are conceptually adopting something called Whitebox AI, meaning it is fully explainable, reproducible in similar conditions, and human certified by all the different work requirements.”
He explains that the approach also helps diversity and inclusion based on the company's direction. He also suggests that the purpose of evidence-based HR is not to find the right answer through prejudices. “We are dealing with people throughout,” he adds. “So the purpose is to use available evidence, research, internal data, analysis, experience, interviews, etc., to find the solution with the highest probability of adding the most value to your organization.”
MapRecruit has heterogeneous data from millions of jobs from hundreds of job boards and abundant resumes. The platform applies artificial intelligence to understand data relations while deriving the ontology with an unsupervised approach. This makes the system continuously learn and create all the patterns based on the data elements. “These algorithms derive attributes which are almost 93% accurate by considering the context while learning from the implicit user feedback,” Sriperumbudur says.
MapRecruit is addressing that newly emerged side of diversity and inclusion concerning people working virtually and remotely, he says. “When reaching out to candidates, you are reaching out in a template of the company. And then we give them all the channels like SMS, WhatsApp, or a chatbot. The chatbot is equipped with 105 languages. Somebody speaking Spanish can ask a question in Spanish. Again, if you are giving a video interview, our platform will help you schedule the date and time via the scheduler based on the availability of the hiring manager or multiple people. So it is in indirectly evoking the softer skills of the person, wherever they are.”
Sriperumbudur further stresses that all organizations have to adopt HR technology to drive transformation, and that, it has to be led by a mindset change.
“The mindset has to be there for the organization. Technology is there to accelerate that transformation. And then, the recruiters have to invoke the HR department saying that, data is the key and then I have to use different tools,” he adds.
He explains that while many companies are already adopting applicant tracking systems but they lack a ‘proactive approach.’ They are driven by reactive approaches like data gathering.
“That is why you really need to know the mission and focus of the organization. That is very important at the leadership level to define and adopt a technology, which can engage with employees or future employees. Using technology just to gather data elements is not going to work. You need to understand new generation tools like MapRecruit which can do more than 80 to 90% of the work because it has all the data elements,” Sriperumbudur adds.
He suggests that organizations need to think about the kind of technology they need to adopt, the current HR strength, and the tech-savviness of existing recruiters. “If the person is not adopting the benefits of the tool, it is like giving you a Ferrari, but you want to go in at 20 miles per hour. So it's a very much a cultural shift within the organization in this entire footprint of the entire line of business functions,” he signs off.