As I continue exploring AI upskilling, I’ve been asking Talent Acquisition leaders how they’re approaching AI.
Not the conference answer.
Not the LinkedIn answer.
The real answer.
So I grabbed David, a TA Operations Leader from Promptmates to ask how he learned to build with AI despite not coming from a traditional engineering background.
Several themes emerged.
Move Toward Change Before It’s Obvious
One of the first things David told me was:
“I am a continuous learner. When I see something moving fast, I move toward it.”
He pointed to the mid-2010s when big data dominated business conversations. Rather than waiting to see where the market went, he enrolled in a Business Analytics MBA because he believed organizations would eventually need people who could understand and use data effectively.
Years later, while working around Robotic Process Automation initiatives, he watched systems begin handling work that had traditionally been performed manually. That experience changed the way he thought about technology and automation.
By the time AI became the hype, he wasn’t trying to decide whether it mattered.
He was trying to figure out what to build.
Start With Friction, Not Technology
One of the first AI projects David built was a tool called Scoutie.
At the time, his recruiting team supported 16 distinct businesses under one operating umbrella. Recruiters relied on an Excel spreadsheet to find information about schools, benefits, and other details candidates regularly asked about.
David looked at that process and as he put it:
“That is not a system. That is a liability.”
So he scraped information from all 16 websites, built a centralized knowledge repository, and turned it into a custom GPT that recruiters could interact with conversationally.
Instead of digging through spreadsheets, recruiters could simply ask questions and get answers in seconds.
He named the tool Scoutie because he believed people would be more likely to use something that felt approachable and useful.
And when I asked about his design process, his answer was straightforward:
“I identify the friction. I ask what a person needs to do, and what is slowing them down. Then I build the simplest thing that removes the obstacle.”
He wasn’t looking for an excuse to use AI. He was looking for a problem worth solving.
That’s a useful reminder for anyone trying to get started. You do not need a groundbreaking idea. You just need a real problem.
Pick One Tool And Learn It
Another point David emphasized was avoiding tool overload.
Many people approach AI by signing up for every platform they hear about and never developing proficiency with any of them.
As he explained:
“You do not need ChatGPT, Claude, Lovable, and Replit running simultaneously. That’s like paying for Netflix, Disney+, Hulu, and Max and watching none of them consistently.”
His recommendation was simple.
Pick one tool. Learn how it works. Build something useful. Then decide what additional tools you actually need.
The Barrier To Entry Is Lower Than Most People Think
One question I hear frequently is whether someone needs budget approval to begin learning AI.
David’s answer was no.
Start with free tools.
He pointed to free tiers offered by major platforms, along with tools like Gemini and Perplexity. He also suggested something many people overlook: search Reddit communities and discussions to see what practitioners are actually using rather than relying solely on vendor marketing.
His perspective on cost was particularly interesting.
“We are in the $5 Uber era of AI.”
I believe his point was that many AI platforms are still offering significant capability at free or relatively low prices. The cost of experimentation has never been lower.
Do Not Wait For Permission
Another strong message from our conversation had little to do with technology.
It was ownership.
David told me:
“Do not let me be the ceiling on your growth.”
Waiting for formal training, enterprise licenses, or organizational mandates can become an excuse for inaction.
His advice was to start where you are. Use mock data. Experiment responsibly.
Build small. Learn continuously.
And most importantly:
“Start building. You are allowed to be a beginner.”
For me, that was the underlying theme of the conversation. I spent months sitting on the sidelines telling myself, “I’m not an engineer, so what do you expect me to build?” That turned out to be the wrong mentality.
The people making progress with AI are not necessarily the most technical people in the room. They’re often the people willing to move toward change, solve real problems, and learn by doing.
Special thanks to David Weinstock from PromptMates for taking the time to answer my questions and share his perspective on AI upskilling, experimentation, and building inside Talent Acquisition.