How AI is making life easier for recruiters
Daxtra Technologies is one of the most innovative RecTech companies in the marketplace.
Founded in 2002 as a spinout from the University of Edinburgh, it has been at the forefront of the AI revolution in the staffing industry, providing recruiters with a suite of automation and productivity-enhancing tools that deliver immediate ROIs.
Today, co-founder Sergei Makhmodov, discusses what he sees as some of the recruitment industry’s most pressing topics.
Q. AI and machine learning capabilities are at the forefront of many industries. How is Daxtra applying these technologies to the recruitment sector?
Recent advances in Artificial Intelligence and Machine Learning have benefited the recruitment industry in several ways. An example is our resume parsing software that mimics the way the human mind reads and understands text.
Daxtra’s software analyses the linguistics structure of free text and applies context-driven reasoning to determine what different words mean in a particular context, and how they are related to other words in the same text passage.
Processing free-form information accurately presents a challenge for computers because words in any language can be ambiguous outside of a context where they are used, and information is often implied rather than being explicitly stated.
For instance, the phrase “Marketing Manager” is the current title of the candidate but only when it is mentioned in the context of the current job in the CV, and is no longer “current” if it appears in any of the past jobs.
At the same time, in the context of “reporting to the Marketing Manager” or “supervising a team of Marketing Managers” that job title has nothing to do with the candidate per se. In order to disambiguate the information that we extract or match, and bring the accuracy practically to human levels, Daxtra developed its own approach to Computerised Text Understanding.
Our technology is based on advanced machine learning algorithms, together with deep learning neural networks. This is used to execute various subtasks of our resume parsing, search, deduplication, matching and ranking.
The levels of automation that we are able to achieve today simply would not have been possible in the past without the recent developments in AI and Natural Language Processing.
Q. To what extent is Daxtra altering the way recruiters find relevant information and manage their databases?
Daxtra automates large chunks of the recruiter's daily routine. The days of spending long hours manually processing and coding candidate’s CVs are long gone. This can now be automated end-to-end at a fraction of the time and cost. This is commonly referred to as “Robotic Process Automation” (RPA) or “Intelligent Process Automation” (IA).
Equally, Daxtra Search delivers impressive efficiency gains for recruiters in the increasingly tough game of “Who puts the right CV on the hiring manager’s desk first”. Such features as “watchdog” auto-searches that automatically deliver matching profiles directly to your inbox even when you are not logged in to your ATS; direct matching from jobs on the ATS (or even from a vacancy URL on your client’s site!), coupled with powerful relevancy ranking – all bring extra efficiencies and a 50-60% reduction on lead times required to come up with a shortlist of suitable candidates.
Data enrichment and keeping your database current has also been a major bottleneck for many decades and can now be fully automated.
We are seeing that 60-70% of the candidates who apply to recruiters’ jobs posted online are, in fact, already on their databases. On one hand this is great because you are getting updated candidate information, but on the other hand, without an accurate automated de-duplication, these candidates just end up as duplicate records on the database.
Candidate data is like a carton of milk – it ages very quickly. And because the data is not routinely updated on the ATS/CRM database, recruiters often don’t even search their in-house talent data, preferring to pay third party job boards to download more recent profiles of the same candidates they already have. This, of course, is a ridiculous waste of effort and money.
This problem can be addressed. After deploying Daxtra solutions, our clients are reporting significant shifts in the numbers of placements that come from their own databases – from as low as 5-10% prior to Daxtra to as high as 60-70% after the automation is in place.
When you do the maths, it is quite staggering how much money can be saved.
Our automation solutions are deeply integrated with our customers’ ATS databases, helping reduce the overall “cost of placement” and shorten lead times to fill roles. Precision, speed, and the broad range of automation features offer our customers a competitive advantage.
Q. Beyond your technological capabilities, what makes Daxtra stand out in APAC and globally?
APAC, specifically, is unique from a languages perspective, being home to 4.5 billion people who speak 2300 languages across 28 linguistic families in total. Luckily for us, nearly 99% of all CVs here are written in just eight languages – Chinese, English, Japanese, Russian, Korean, Vietnamese, Malay/Indonesian and Thai.
So, accordingly, we focused our efforts on those languages and made sure our linguistic software can handle them well. As we process more data in those languages, the software gets more and more accurate, it is a continuous learning project.
The new big thing that we are working on is better cross-language matching – if a job description is written in Chinese, for example, you do not want to miss out on suitable candidates who submitted their CVs in English.
This is now available for Japanese (to English and back), Chinese and several European
Q. As global events cause fluctuations and shifts in the job market, is Daxtra set up to deal with an ever-changing environment?
With the current disruption due to COVID-19, we are seeing some immediate and substantial changes. Companies most severely affected by the downturn are understandably cautious about spending money, but even those with fewer jobs to fill are most likely handling increased volumes of applications lately.
In any such crisis, the situation flips. Now, we are in a situation where we have an abundance of resumes, candidates and jobseekers but not enough vacant jobs.
The benefit of staying in touch with your talent community during such a period is that candidates will, later, be happier to engage with a company that has shown it cares - whilst keeping in mind that some of those candidates may one day become clients.
It is important to try and capitalise on this. The question is, how do you do this if your own workforce has shrunk? That’s where Intelligent Automation (IA) can help.
Taking a broader view on the near to mid-term future, for the most part, the industry is still on a clear path towards further consolidation in both the Recruitment and the RecTech sectors.
Also, as more corporate organisations are bringing Talent Acquisition teams in-house, we will be seeing increased competition between recruitment agencies and their own clients.
In the RecTech space most notably, several relatively new technologies are gaining strongholds in the current shrinking economy. AI-based chatbots and candidate engagement tools are likely to grow significantly in popularity, along with “autonomous staffing”.
The traditional “bricks and mortar” agency model will start giving way to an automated approach as more hi-tech staffing firms are increasingly filling positions with minimal human involvement. AI and IA are progressing at such a rapid pace that more and more positions will be filled autonomously in the coming few years.