AI and more specifically Large Language Models (LLMs) are changing the world of recruitment.

Recruitment has already become exponentially more data driven than it used to be.

Billions of candidate profiles are to be found across a variety of platforms and every profile has become more data rich than ever with people becoming more active on social media like LinkedIn, in forums like Reddit and professional open source websites like GitHub.

To be able to do something with this vast amount of data you cannot rely anymore on your human abilities only; it’s impossible to shift through a billion profiles one by one. That’s why we’re already using AI systems in recruitment like search engines which deploy AI to get from millions of profiles to just your top 100 candidates.

But the AI systems that are being used are becoming a lot more sophisticated with the more recent introduction of LLM’s like OpenAI GPT (GPT 3.5 and GPT4) but also a wide variety of other LLM’s like Google’s LaMDA + Bard and open source LLMs.

These new AI systems also allow for ‘Generative AI Recruitment’, which means they can generate content on top of being able to analyse them like conventional AI systems. 

Next to the fact that AI has already been proving itself as more than just a trend, the practical value for this technology, in the recruitment industry specifically, has also become a lot more clear.

The recent introduction of language models is particularly interesting for the recruitment industry.

Recruitment comes down to basically the processing of language.

Think about your daily activities as a recruiter:w

• Writing a job description
• Searching for certain job titles and keywords
• Looking at what a candidate has written on their profile
• Communicating with candidates

Language is the primary driver of recruitment activities and hiring decisions.

Recruitment = the interpretation of language and the enticement through language.

Being aware of the role that language plays in the recruitment profession is the start of better understanding the impact of language models in recruitment.

Examples of GPT powered recruitment solutions

Although AI language models can be powerful, they mean nothing without the right data. If you let GPT for example work on a limited database, you can also expect limited output. The size and relevancy of the data of the AI recruitment software are key in determining its usefulness for you.

Here are some examples of AI recruitment solutions in the market and their features:

AI Recruiter (Uwi)

AI Recruiter Uwi is a completely autonomous AI recruitment assistant. Uwi recruits talent for your job tirelessly by completely automatically search, screen and reach out to potential candidates based on your requirements and her general knowledge from the web.

Data: 1 billion profiles worldwide across a variety of platforms like LinkedIn, GitHub, Stack Overflow

Outreach: next to searching and screening, outreach is also completely automated with hyper-personalized prompted messaging and auto-follow up

RecruitGPT

RecruitGPT is a recruitment solution built by HeroHunt.ai. RecruitGPT can create a shortlist of candidates from just a single line description of who you’re looking for. You ask RecruitGPT for example “Hey RecruitGPT, find me a full-time java developer with Python, C# and API skills in Amsterdam” and it will give you the list with top matches in seconds.

Data: 1 billion profiles worldwide across a variety of platforms like LinkedIn, GitHub, Stack Overflow

Outreach: completely automated with messaging on autopilot and auto follow ups with verified personal email addresses

Seekout Assist

Seekout is a profile database. It allows recruiters to find hard-to-find talent with boolean search, but recently also with a more automated way of searching based on language models. Seekout also indexes profiles from GitHub, research papers, and patents. 

Data: 700 million profiles worldwide

Outreach: partly automated outreach with email (not always verified)

PeopleGPT

PeopleGPT is a very early stage recruitment solution based on the GPT language model with a database of primarily US profiles. The idea of PeopleGPT is that users can find talent based on a chat interface.

Data: database with a selection of LinkedIn profiles limited to US

Outreach: semi-automated with templates but one on one sending

How to use AI systems in your daily work

You already use AI in your recruitment activities.

The platforms and search engines that you use to find talent already use AI to determine the best results, even if you do a simple search with for example only a job title on LinkedIn AI systems are already involved.

So the question is not if you should use AI, because you already do, but the question is how you can differentiate yourself by using AI and the newer language models in a smarter way.

These are key recommendations to better understand and use AI:

1. Follow the developments in AI
2. Trial different tools
3. Make decision based on usefulness, not newness
4. Look for a ‘full cycle’ solution
5. Scale your operations with the possibilities op AI

Follow the developments in AI

In order to make use of AI in the right way you need to understand where AI solutions in recruitment are now and keep track of the recent developments. AI solutions will develop exponentially quicker, which means that in two years from now these systems have become a lot better than they did in the two years before that.

Since the speed of technological advancement is higher than it used to be, it’s more important now to stay up to date with today’s leading solutions and developments.

Some recommendations to follow: Recruiting Brainfood, HeroHunt.ai blog 

Trial different tools

Go for a trial period or at least a monthly subscription before you buy an AI recruitment solution for the longer term. Generally speaking, a demo for an AI tool is not sufficient to show its value. You have to get a feel for how the AI tool behaves for your specific use case. 

Recommended is to request a free trial period for each solution that you’re reviewing or go for a monthly plan initially before committing to longer term contracts.  

Make decision based on usefulness, not newness

There is a lot of hype currently around AI. This creates a lot of noise of trending AI tools which seem exciting but are actually not that useful. 

Don’t get dragged along with the hype but rather do your own research to what you think would work in your current recruitment activities. Think of that one tool that you are missing right now and start looking for that proactively.

That way you can make sure you make decisions based on the usefulness to achieve your recruitment goals instead of being distracted by overhyped tools.

Look for a ‘full cycle’ solution

AI is becoming more and more generic. Where AI solutions used to serve one very narrow purpose like search, AI solutions currently are heading to more generic solutions which can serve an entire professional domain.

This means that there are more and more recruitment solutions out there which increasingly serve as a full cycle software solution instead of a point solution.

You don’t need to tie 10 different tools together anymore (search engine, contact finder, messaging tool, analytics…), you can do all of those things with one and the same tool. 

Scale your operations with the possibilities op AI

The big promise of AI solutions is far-reaching automation. If you are able to increasingly automate your recruitment activities you can also further scale your recruitment activities. You can do a lot more with less time and less people. The key here is to think bigger in terms of scale (eg how many people you can reach out to daily) and act accordingly.

Set a challenging and clear goal for yourself and select new tooling based on that.