I sat down with a TA leader from PromptMates whose career has evolved from agency recruiting into helping early-stage companies build hiring functions, recruiting operations, and AI-enabled recruiting workflows. Since much of his work revolves around building pipeline from scratch, I asked him a simple question:
If you walked back into Robert Half tomorrow, were handed a cold territory, and had access to today’s AI tools, what are the first three things you’d do?
Here’s what he told me:
The first workflow would monitor his local market for new job postings that match his target accounts. Instead of manually checking job boards every morning, an agent would surface hiring activity as it happens. For someone building a territory from scratch, knowing who’s hiring today is far more valuable than finding out next week.
The second workflow would review his outbound messaging. Rather than simply generating emails, it would analyze outreach sequences, identify weak points, and recommend improvements. Most recruiters spend time using AI for initial outreach or still do it manually. Far fewer spend time evaluating whether their messaging is actually getting better over time.
The third workflow would run while he sleeps. Every night it would identify new prospects, enrich their profiles, and generate personalized outreach based on the company’s hiring activity and likely business challenges. Instead of starting each morning with a blank slate, he’d start with qualified prospects ready for review.
As our conversation continued, another theme started to emerge: Every workflow he described was designed to eliminate repetitive work so recruiters could spend more time solving hiring problems and building relationships.
He gave me an example from his own work. Many early-stage companies don’t have an applicant tracking system when they begin hiring. Recruiting often starts in Google Sheets, with recruiters manually copying candidate information, linking LinkedIn profiles, updating stages, and maintaining pipeline reports.
After spending just a couple of days doing that work, he built an MCP server that connects Claude to those spreadsheets. Instead of manually updating candidate stages, he can tell Claude to move a candidate from recruiter screen to hiring manager interview. He can ask for a pipeline report showing where candidates are exiting the process. The administrative work happens through natural language instead of repetitive clicks and copy-and-paste.
That conversation reinforced something I’ve been thinking about as I continue talking with recruiting leaders who are actively building with AI. The discussion is becoming less about finding new ways to automate recruiting and more about identifying the work that never required a SME in the first place.
Building relationships, understanding hiring managers, evaluating talent, and earning trust requires people. Updating spreadsheets, moving candidates between stages, and assembling reports do not.
One thing I’ve realized after talking with staffing leaders over the last year is that the people who describe themselves as “old school” are often the ones who would benefit the most from this. When I think of old school recruiting, I don’t think about manually updating spreadsheets, reviewing hundreds of resumes, or clicking candidates from one pipeline stage to the next. I think about being on the phones, meeting clients, building relationships, and developing business. The top operators I’ve met aren’t using AI because they want to recruit less. They’re using it to cut down on administrative work, simplify their tech stack, and create more time for the parts of the job that actually generate results. In many ways, that feels like a return to old school recruiting, not a departure from it.
A big thank you to John LaMarche from PromptMates for taking the time to share his perspective. Conversations like this are one of the reasons I enjoy being part of the PromptMates community. They’re practical discussions from people building real recruiting workflows, and every conversation leaves me with a few more ideas worth exploring.