Abhinav Agrawal is the CEO of Rocket, an AI-enhanced recruiting agency as well as Hireflow.ai, a free outbound recruiting solution.
The pandemic has inspired fear of the incentive to replace employees with machines. As the co-founder of an AI-powered recruiting organization, I expect that AI won’t displace recruiting jobs. In fact, I anticipate that AI will become an indispensable tool for recruiters.
Ask any recruiter: What’s the most tedious aspect of your job? They’ll likely tell you it’s looking through piles of resumes in order to decide who is the right fit for a job. Over the years, there have been technological developments to improve this process (e.g., keyword matching). However, today’s algorithms are much more sophisticated, incorporating a wide range of attributes about a candidate and using complex machine learning (ML) models to map out job hierarchies. They’re capable of telling the difference between a front-end full-stack engineer and a back-end engineering role, and more. AI and ML allow profiles to be categorized quickly and accurately so that you’re not left second-guessing the results.
AI streamlines work processes so that HR managers can concentrate on more “human” skills, including relationship-building, interviews and making the perfect placement. For candidates, the need is to find the next, safe position. AI can automate much of the busywork so that tasks such as vetting prospects, reading emails, drafting responses, managing calendars and updating the ATS don’t subtract from the key work of matching the right person to the right job.
Job searching is a lot like dating — everyone’s looking for the right match. For recruiters, how do you help people find what is right for them? For individuals, how do you know what is right for you? A mentor can help individuals, but that’s difficult to find when in-person meetings are rare. What if we used data to help answer these questions?
Say a candidate is in the beginning stages of their career, they work for a large, public company in a fixed role. What’s the next, best career move? If we use data to help with that decision, we can organize factors in various buckets: size, funding, valuation, salary, career acceleration, mentorship, etc. These factors can be rated according to relevant value to the candidate, indicating whether or not a job possibility is a solid choice.
Making a hiring or career decision is hard. But determining better career trajectories is an example of how big data and AI have an important role to play in candidate selection.
Significant recent breakthroughs in ML allow technology to be effective for recruiting. Self-learning algorithms are able to evaluate troves of data the way a human might, but better. These advances allow AI to quickly and accurately help human recruiters organize and prioritize tasks so they can spend more time on what matters — finding the best role.
Recent civil society movements have made it clear — it’s critical we become conscious of our biases and make decisions to eliminate them.
Building a better recruiting algorithm can help. A good algorithm can focus not on the associative traits that informed the search, but on the skills that a person has. A good algorithm, though, removes variables you believe are not related to a job.
This is where AI can help humans. Whether we intend it or not, we have unconscious and conscious biases. When it comes to recruiting, biases can lead to mental shortcuts for candidate selection. Some jobs, because of historical legacy issues (e.g. software engineering) may be more white and male. But if software took a harder look, focusing on diversity, it may be easier to match with applicants from a variety of backgrounds.
No discussion of AI’s merits is complete without considering the challenges or limitations it presents. In the field of statistics, for example, bias is acknowledged as omnipresent to the extent that there are designations of types: sampling errors, measurement errors, biased estimators, etc. Bias can enter through the collection of data — in the questions that are asked, for example — or in the constraints set up when creating a survey. Machine learning finds patterns in data, so if a data set is biased the AI system will be too.
The best first step for business leaders who are considering using AI-based recruitment is to acknowledge that bias exists. Curbing prejudicial bias in an AI-recruiting system means asking questions about how a data set is formed and what steps were taken to question bias. Also required is regular monitoring of the devices used to create data; does a camera use a chromatic filter that may create color bias? Each solution must be specifically tailored to an organization’s specific process.
Three Tips To Get Started
1. Understand your existing system. Examine your company’s recruiting process and identify areas that need improvement. Is it sourcing? Screening? The application process? These are all areas that can potentially be improved via AI.
2. Consider hiring volume. Is your organization hiring hundreds or thousands of people? Are resumes pouring in? AI can evaluate thousands of applications a day, quickly determining if a candidate meets the job requirements, trimming the eligible pool so it’s more manageable for a human recruiter. Machine learning is also adept at sorting through great amounts of data.
3. How typical is the role? A UX/UI designer is a common position in many companies today, and an AI system has likely searched for it thousands of times. As with any data set, more information increases accuracy. It’s more likely that an AI system will recognize if a candidate’s characteristics are relevant if it’s been trained on the same search many times before.
AI has the potential to become an indispensable aid to recruiters with its capabilities to streamline candidate sourcing, reduce busywork, and even work to diminish bias in the hiring process. AI, and its self-learning algorithms, are here to help recruiters focus on the “human” in human resources.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?