21 days, $5K, 7 AI agents: how a non-programmer built a talent marketplace

How 10 Years Trying to Improve Executive Search Finally Clicked When I Locked Myself in My Office for 21 Days With 7 AI Agents and Built Something for You That Doesn't Exist Anywhere Else.

Strategic Networking

I need to tell you how this happened. Not the polished version. The real one.

The short version is that my partner Mike Martin and I just launched the Bearhug Network, a two-sided marketplace that connects the people who hire executives with the executives who are either actively looking or are open if the right opportunity were to come knocking.

Anonymized profiles. Vetted by the Bearhug team. Browse, filter, request an introduction, and we handle everything from there.

If you've ever been asked "who do you know" for a critical hire and wished you had a better answer, or if you've ever wanted to quietly explore what's out there without broadcasting it to the world, this is what we built and why.

The longer version is that this thing has been living in my head for a decade. I tried to build it twice before and stopped short due to the enormous effort and expense, and then it accidentally came into existence over 21 sleepless days in May 2026 as a side effect of a completely different project. Roughly 350 hours managing 7 coding agents in tandem (a very different experience than managing humans, I must say), about $5,000 in total investment, and somewhere north of 75,000 lines of production code later, here we are. I've never written a line of code in my life, and do NOT consider myself technical (at all).

But I need to start at the beginning.

Why This World Captivated Me

From the very first time I learned about executive search, I was hooked. A small niche cottage industry built on trust and relationships, where a service provider like me could be handsomely compensated for the value of helping make an executive placement. Obviously intriguing. But the thing that was far more important was the actual work itself.

My entire career as an entrepreneur since the age of 19 had been in marketing, business development, and sales. Not executive search. But the work hit the bullseye for me in a way that checked nearly every box for my personal interests and genetic wiring.

I've always been fascinated by why businesses succeed or fail, how teams are built, how products find product-market fit, how capital is raised and managed, and how ultimately business can be a positive force for good.

In executive search, I saw a chance to insert myself into the most interesting conversations at the board level and play a key role at a pivotal moment where the right hire could determine whether the business reached its next stage of growth.

This is why "Topgrading" by Brad Smart sat at the top of my favorite books list long before I knew executive search was even a thing, and why "Who" by Geoff Smart (Brad's son) dethroned it once I got into the business. It was validating to find the subject matter and see my place in the world through that lens. That validation gave me the tenacity to jump in without any formal training or prior search experience.

Every top performer I interviewed before starting Bearhug told me there was zero chance anyone could succeed without first joining one of the top five firms for at least five years. Yet despite everyone saying I'd fail, here I am still standing a decade later, and the lessons from that decade are baked into every feature of the network you're about to see.

10 Years, 18 Unique Experiments, and the Lessons That Led Here

Having come in with a fresh perspective rather than inheriting the established playbook, I questioned everything. The way searches are conducted. The way models are commercially structured. The way candidates are screened and managed through the process. And how to make use of all the incredible people I'd meet during a search who wouldn't make it to the presentation stage, people often discarded with nothing to offer them.

Not to mention the thousands who come inbound every year hoping we can help them land a job, despite our job being to find people for jobs, not jobs for people, as we headhunters say behind the scenes. That reality pains every one of us. If you're one of those people, this story is especially for you, because the network we built is designed to finally give you a seat at a table that's been invisible until now.

Over the past decade, I've attempted no fewer than 18 distinct models to shape and evolve how executive search is conducted. I'll name three.

CEO Flow was a subscription recruiting model for early-stage venture-backed CEOs. Instead of a one-time placement fee, CEOs would pay monthly and get multiple fully-managed candidate pipelines with no placement fees attached. Reduce their risk, increase their capacity, create a continuity-based partnership rather than a reactive one-off engagement. We sold it. But the convention of how search is bought and sold was too ingrained for broad adoption.

Top of Funnel Talent let CEOs share responsibility for managing the search lifecycle to dramatically reduce their fee. We'd handle the first half (strategy, sourcing, screening, shortlist), then hand off to the client to run interviews and close. The big idea was that 80% of the expertise lives in the first 20% of the search lifecycle, so by handing off at the halfway point with a vetted shortlist, we could cut fees by more than half.

It was a mess. The candidate experience suffered because clients didn't pick up the handoff with the same white-glove treatment. Leaving them to close without us brokering in the middle left too much to chance. It only reestablished why the full end-to-end process matters. The concept eventually evolved into a fully automated version for mid-level recruiting that works well, but for executive search, the lesson was clear: the full process is worth the full fee.

The Fund-Level Human Capital System was built for venture and PE firms. Instead of solving leadership gaps one portfolio company at a time after problems surface, we proposed a proactive system with a two-way sync between a fund's upcoming needs and our ability to pre-recruit on behalf of the entire portfolio. Each month, portfolio CEOs would receive a talent digest of 100 to 1,000 pre-recruited executives who'd already raised their hands to take exploratory calls, complete with contact info, work history, and an automated booking system.

The system is still live. But it requires funds to buy into a paradigm they haven't seen before, and that takes time. For those where it fits, imagine being a portfolio CEO who can submit upcoming needs and have pre-vetted talent delivered on a silver platter the next month.

What these three and the other 15 experiments I ran had in common was that they were too unconventional for wider adoption. Not because they were wrong. Because they didn't meet the market where it was. After a decade of trying, I realized the better path is to meet the market where they are, misconceptions and all, while still delivering the most incredible experience possible for the most value. That realization is what led to the Bearhug Network, and why the network is designed around how you already think about hiring and career moves rather than asking you to learn a new paradigm.

What Buyers of Search Actually Want (and What They Get Wrong)

Talking with thousands of board members, investors, and CEOs evaluating talent partners, I've learned three things. If you've ever hired a search firm or considered it, you'll probably recognize yourself in at least one of these.

First, they want proof of past success doing searches as close to their immediate need as possible.

This makes sense. "Yes, I just did three of those exact searches. I know all the people. I can succeed with yours too." Sure.

But search is search, at least the way we do it. We start from scratch every time with a beginner's mind, build the candidate list from the ground up, and in 45 to 60 days we've got a signed offer letter. Firms that specialize in one function are more incentivized to recycle people they already know. We always start from scratch. It takes an extra week, but the coverage, depth, and audit trail give us a much higher probability of finding the best right person. And due to the other efficiencies we bake in, we're 2.5x faster than the industry average.

No disrespect to headhunters who already know the eight people they're going to call by the time they arrive back home after an intake. That's powerful. But is that really the best way to serve your client? That's a debate I'll never stand down from.

Second, they find comfort in going with an established, well-known firm.

But it's not the logo that matters. It's who is actually executing the search. The bigger firms chop up work and delegate key parts to junior staff. Searches take longer. And because those firms are typically paid their full fee by midway through the search, motivation fades when things get hard.

Boutique firms like Bearhug get a lot of work cleaning up the aftermath. By going boutique, clients also get the benefit of our lower overhead equating to them getting a better deal, a higher-quality experience, and searches that close in about half the time.

Third, almost every buyer of search has asked us for one-off freebies.

"If you run into any great CMOs, will you let me know? I've got a friend who's hiring but isn't ready for a search yet."

Or the classic: "I've got a CFO friend who's a rockstar who's active in the market. Can you help get her a job?"

That last one happens constantly and stings a bit every time. Why? Because that's simply not how this industry works. The chances of randomly stumbling upon the right person (or the right job for a person who's looking) with the hundreds of little things that magically have to line up is sub-1%.

We're always happy to try and help, but there's only so much matchmaking you can justify doing for free, aside from the fact that without doing at least an hour-long intake with the hiring authority, it would be worse than trying to throw spaghetti at the wall hoping something might stick. And that's not the business we're in.

This was actually the primary impetus for the Bearhug Network. If you've ever made either of those asks (and most of us have), the network is the answer you were looking for. Browse anonymized profiles of vetted executives, filtered by function and expertise, and request an introduction when someone catches your eye. No search engagement required. No fee unless a placement is made.

The Accidental 21-Day Sprint

Building a two-sided marketplace for executive search had stalled out on me twice before. A few hundred grand to build, a team of engineers, 6 to 12 months of work. Too enormous. But I couldn't let it go.

Then May 2026 changed everything. We all remember the moment around the turn of the year when the headlines hit that developers at the biggest tech companies were no longer writing their own code. It was being written by the model itself. I knew my time would come, but I stayed on the sidelines waiting for the right moment.

In early March, I'd built a personalized email campaign using Claude AI and started doing technical things I'd never tried. By late April, I had an ah-ha moment. For 18 months I'd thought we were "AI native" because we were using AI tools across our workflows. But I hadn't grasped what truly AI native means.

For us, it meant operating all our workflows through prompts run by Claude. So I began creating what became known as "the brain," with Claude as the intelligence layer orchestrating everything through the apps it connected to. We fast-tracked getting 100% of our data out of disconnected silos, dumped it into the brain, and killed off any tool that didn't have a deep integration. That realization unexpectedly kicked off a 21-day sprint that consumed my entire life. Eighteen-hour days straight.

What I wasn't

[truncated for AI cost control]