One of my favorite parts about interviewing Talent Acquisition folks inside the PromptMates community is realizing how often I walk into a conversation expecting one thing and leave thinking about something completely different.
I thought my conversation with Jess was going to be about Claude Skills, Plugins, and Change Management.
And I mean…it was.
But what continues to stand out in these convos isn’t just the technology…it’s the way my peers are approaching problems.
Literally every example she shared started the same way:
“There has to be a better way…”
Build for a problem, not for AI
One of my favorite examples was a Claude Skill she built called the Try-Out Task Builder.
At Checkatrade, candidates complete a take-home assignment as part of the interview process. Jess explained that hiring managers were creating these tasks in completely different formats. Some were several pages long, others were only a few sentences, and the quality varied depending on who was hiring.
Rather than asking every hiring manager to change how they worked, she built a Claude Skill that takes whatever brief they provide and turns it into a consistent, candidate-facing try-out task. It also generates separate reviewer notes so interviewers know what they’re looking for when evaluating submissions. If a brief is too vague, the Skill asks follow-up questions before generating anything. If the expected completion time seems unreasonable, it flags that too.
The result is a more consistent candidate experience, less back-and-forth with hiring managers, and a process the TA team owns instead of recreating for every new role.
She told me she built it in Cowork simply by describing what she wanted to achieve. More importantly, she wanted the entire TA team to be able to use it themselves instead of relying on her every time.
This was probably my favorite example because it immediately made me think about the candidate experience.
If organizations are going to continue using take-home assessments, consistency matters. Candidates shouldn’t have completely different experiences simply because they’re interviewing with different hiring managers. By standardizing the assignment while still allowing hiring managers to provide the input, the TA team can create a more consistent experience without adding more manual work.
Turning reports into actions
Another build example that Jess walked me through Tabitha, a plugin she built around Ashby data.
Instead of simply displaying pipeline metrics, Tabitha identifies stalled roles, scheduling bottlenecks, candidates sitting too long in a stage, and roles that appear active but have little momentum.
It also tracks KPIs against targets using RAG status and prepares talking points for hiring manager conversations, ready-to-send follow-up messages, and even a Wednesday recap designed around the team’s weekly meeting.
What I appreciated most was how different people receive different views.
Talent Partners see their own pipelines.
The Senior Talent Lead sees both personal and team-wide activity.
Jess has a TA Operations view across the organization.
There are separate executive views, and she’s currently building one for the C-suite.
Everyone receives the information that’s relevant to them instead of sorting through reports that weren’t built for them.
Building through iteration
Another Skill she showed me was called The Interrogator.
It didn’t actually start as The Interrogator.
She originally wanted a personal workspace where she could capture notes and ideas. Over time, that evolved into something that interviews her own plans by asking questions about assumptions, dependencies, stakeholders, risks, and alternative outcomes before she moves forward.
She described it as a way to stress-test ideas rather than trying to make the first version perfect.
That iterative approach came up several times throughout our conversation. She didn’t describe sitting down with a complete blueprint. She started with a problem, built something useful, continued using it herself, and refined it over time.
Her advice for getting started
I asked Jess what advice she’d give recruiters who haven’t really started using AI yet.
Her answer was simple.
Start talking to it.
She shared that one of her earliest use cases wasn’t building applications or creating complex workflows. She simply used AI as a sounding board.
From there, she built a new starter onboarding app with Lovable. AI helped her organize her thinking, structure a plan, improve her prompts, and eventually bring the project to life before partnering with an engineer on privacy considerations.
Her advice was that you won’t know what’s possible until you start asking questions.
She also emphasized something I think is worth repeating: AI isn’t always right. Human oversight matters, and the more you use these tools, the more they learn how you work and communicate.
Adoption doesn’t happen overnight
Our conversation also shifted toward something every leader faces once they build something useful.
How do you actually get people to use it?
Jess said everyone is at a different stage with AI, and that’s completely normal. Some people are naturally curious. Others are more cautious. Rather than expecting everyone to immediately embrace a new tool, she tries to meet people where they are.
For example, when she rolled out Tabitha, the team walked through it together, but they were also given time to explore it themselves. Having support from the Senior Talent Lead helped reinforce adoption, but she also stressed that people need to understand why a tool exists before they’ll incorporate it into their workflow.
The other piece she kept coming back to was feedback.
Release something. Ask people how it’s working. Improve it based on what they tell you. That feedback loop isn’t optional.
Final thoughts
I came into this conversation expecting to learn about Claude Skills.
I left with pages of notes about solving problems, iterating, and helping teams adopt new ways of working.
Huge thank you to Jess Martin from PromptMates for taking the time to walk me through what she’s building at Checkatrade and for sharing both the successes and the lessons she’s learned along the way.