top of page

I, Recruiter (part deux) – AI and the process of recruiting talent

Being human

A recent blog I posted entitled ‘I, Recruiter’  generated interesting responses from individuals from the recruitment industry and senior level tech professionals as well. Thanks to all for their contributions to the wider debate regarding AI and recruitment.

Expanding the debate further, I am seeing more evidence of ‘bot solutions’; a colleague of mine (*thanks James) shared an article this morning about Westpac’s intention to use “bots” to respond to customers simple questions through their current digital banking channels. If banking clients are prepared to trust AI to improve customer interactivity then surely this could be considered for any customer-facing industry such as tourism, gaming/betting, retail and recruitment?

Some might argue that Westpac’s decision to extract the ‘human element’ will result in a less satisfactory customer experience. I would counter by saying that the customer’s experience may actually improve as the quality of the service for more complicated or complex queries are handled by a human that has more time to dedicate to the individual customer.

Changing behaviours

Some of the commentators on my original post mentioned that the relationship between recruitment consultant and candidate will not be replaced by bots or technology; in fairness, I’d say that’s a fair statement (*additionally, I never said they would). What I do believe is that the search-&-selection process will be the first part of the recruitment lifecycle to adopt AI more broadly and this will progress to the assessment and interview process; as evidenced by the services offered by Wipro Holmes. So in the latter context, this is already a reality. Now to the search process — has anyone heard of Word2vec or Doc2vec? — they would be classed as machine learning and data mining under the broader banner of AI.


Word2vec is a deep learning technique that allows a model to be built for a particular dataset in which words are used as vectors. This was first bought to my attention by the research and development that has already been conducted on word vector propositions (‘word2vec’) in the 2013 paper written by Tomas Mikolov et al and his colleagues at Google. The principle aim of word2vec is design algorithms (*and then software) to understand the relationships between words with no human guidance. This happens through the application of ‘representation’, combining ‘syntax’ and ‘semantics’ and ‘vectors’.


Doc2vec modifies the word2vec algorithm to a more unsupervised type of learning of continuous representations for larger blocks of text, such as sentences and paragraphs in entire documents.

Changing search

By utilising the resulting software created on these algorithms we could see a reality where internal HR and recruitment functions are regularly using these tools to find talent. Imagine being able to feed in the CV of your top performing developer, UX/UI creative, data scientist and receiving either a profile/CV of an individual that has been highlighted as having the skills and experience of your top performer; and this has all been done through monitoring social media channels and pre-existing information on business social media sites such as LinkedIn or employment websites like HIRED or Monster.


What does this specifically mean for the agency recruitment consultant? — it’s a positive as it will mean that she/he has more time to dedicate to building the relationships that are vital to mutually beneficial business partnerships. I think there may have been a misconception with my previous post that I, in some way, believe that recruiters will be replaced by robots; this couldn’t be further from the truth.

My aim in both the last post and this one is to illustrate that all industries will be disrupted by the advantages delivered by AI and digital technologies; the recruitment industry will not escape this either. The companies that will sustain their longevity and growth are the one’s that understand the benefits of AI, are able to harness it to offer their people new ways to improve efficiency, increase interactivity and offer more value to their clients and business partners.

💡 References

The Sydney Morning Herald — “Westpac puts robots on trial in digital banking platform” — Clancy Yeates — 19th July 2016

Distributed Representations of Words and Phrases and their Compositionally by T. Mikolov, I. Sutskever, K. Chen, G. Corrado and J. Dean — 2013



bottom of page