From Poached Eggs to AI Recruitment Agents
Two years ago, if you had asked me how comfortable I was with building systems from scratch, writing code, creating automations or using AI to support recruitment, I would have said, ‘not comfortable at all’.
My experience with AI back then was mostly using ChatGPT. I would ask trivial questions like, ‘How many penguins would it take to defeat a polar bear?’ (Turns out, not as many as I thought.) I also used it to work out how not to over-poach an egg.
Fast forward to today, and I am confidently using Claude, ChatGPT and several other AI tools to make both my work and parts of my personal life genuinely easier.
What I Have Actually Built
Last year, I joined Street Group, a proptech business that had already embraced AI across both its products and internal ways of working.
By early January 2026, it had become clear that AI was not just for engineers. It was capable of far more than helping a poorly prepared chef overcome subpar cooking techniques.
With Claude already available across the business, I decided to jump in with both feet and see what I could build.
For the first couple of weeks, I spent hours researching, watching YouTube videos, testing things, breaking things and generally trying to work out what everyone else was doing.
Once I realised that there was not exactly a “rule book” to this, everything changed. Instead of trying to copy what everyone else was building, I started looking at where my own bottlenecks were.
For me, it was all the administration that comes with recruiting in a scaling tech business.
Writing job adverts. Drafting outreach messages. Creating interview questions. Revisiting silver-medallist candidates. Researching competitors, for talent.
Rather than trying to build one enormous solution from day one, I started by creating a series of standalone tools. Each solved a specific problem. As I continued iterating, it became obvious they should all work together.
That eventually evolved into my own AI Recruitment Agent.
From a few kick-off notes, it can now generate on-brand job descriptions, define what great looks like for a role, draft LinkedIn outreach messages, create a LinkedIn post and even search our ATS for candidates who might already be a strong fit.
Work that used to take a couple of hours now takes less than four minutes.
I have also built a response assistant that reviews my job board messages each day and drafts replies based on previous conversations. Nothing is ever sent without my review and approval, but every week the quality of those drafts gets a little better.
Could I have bought tools that already do parts of this?
Get Will Small’s stories in your inbox
Join Medium for free to get updates from this writer.
Probably.
But building them around the way I actually work has been a genuine game-changer.
It certainly has not all gone to plan.
At one point, my AI decided it was perfectly capable of messaging candidates without asking me first.
Thankfully, the message itself was absolutely fine. The fact it sent it was not.
That reinforced something I already believed. Recruitment is, and always will be, about people. AI can help me work faster, but it will never replace genuine human conversations.
Why Getting Ahead of the Curve Matters
I completely understand why people have reservations about AI and what it means for the future of work. Those concerns are real.
But AI is not something that is coming. It is already here, and it is already changing the way we work.
I certainly did not do this on my own. I am incredibly lucky to work for a forward-thinking business that invests in its people and gives us the freedom to experiment. Being surrounded by colleagues who were also building their own AI tools definitely accelerated my learning.
For me, the best thing I could have done was lean into it. Learn it. Break things. Build things. Figure out what worked and what did not, while making sure the judgement, empathy, relationships and final decisions always remained human.
These tools now save me hours every week on administrative tasks, giving me more time to focus on the parts of recruitment that will always need a human: speaking to people, building relationships and making great hiring decisions.
The biggest lesson I have learned is not how to build AI tools.
It is that you no longer need to be an engineer to build genuinely useful things.
Twelve months ago, I would never have believed I could build any of this.
Now I cannot wait to see what comes next.
If you are on a similar AI journey, I would genuinely love to hear what you are building.
And if you are curious about anything I have created or would like a hand getting started, feel free to message me.
To find out more about life at Street Group, follow us on LinkedIn, see what our team are saying on Glassdoor, or visit our careers site.
