Digging for gold: how machine learning can disrupt the IT talent acquisition market

Talent acquisition and retention is the single most determinant factor of success in IT companies. It’s also the most cumbersome, unreliable and stupid process I ever experienced. Being on both sides of the fence, I got astonished by the lack of preparation of candidates applying for jobs, fuzzy job descriptions and lack of consistent evaluation metrics and processes in companies and biased reviewing process of candidates. Overall, I got surprised by the lack on innovation in this sector. It seems that recruiting top talent hasn’t change for the last 100 years.

I start analysing why such a critical process is still so much awkward, time-consuming and inefficient. Being a data scientist, I focused in framing solutions based on machine learning and big data that could transform a process based on gut feelings and impressions to a data driven one. This is a very complex process heavily relying on human judgments. But I’m convinced that, with a wise usage of a myriad of data points, from companies and from candidates (either passive or active), we can dramatically improve the efficiency of talent acquisition.

Most companies haven’t fully realized how fast the IT recruitment market has changed. They are still attached to the demand surplus paradigm: “there are plenty of good candidates fighting for this job offer. All we have to do is weed out the weak and select the best”. But rules of the game have changed since A: there is no surplus of talent and B: the best will not necessarily apply.

But perhaps the worst mistake made by companies is, what I call, skills tick boxing: “We post a job offer with a set of desired technical skills and select the candidates with the highest number of proven capabilities”. This is wrong for several reasons: 1) Skills doesn’t mean performance, 2) Skills say nothing about the potential and 3) No “soft” skills are benchmarked. Reality is that the majority of bad hires the problem lies not in the technical skills but in a soft skills mismatch. This include, lack of self-motivation, leadership, cultural misfit, lack of commitment or initiative. If your company have open technical positions for several months, or even years, maybe it’s time to rethink your strategy.

The project we are working is a platform to help organizations identify, target, screen and match good candidates. It relies on the premise that search should be active and we should collect all data points to create a 360 profile of individuals. Most important, competencies are assessed not only on the hard skills (which is the easy problem) but on the soft skills. Relying on a data intensive, the end result is a short list of potential good hires for each company and the reasons for that.

We are aware that hiring cannot be fully automated and inevitably require some human interaction. But we are committed to reduce this effort to a bare minimum relying on technology like deep learning and NLP and removing human bias and intuition with accurate and reliable metrics. Machines and big data will give a much better and sound base to disrupt this industry.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.