What I Told Stanford MBA Students About Finding Product-Market Fit

Lessons from the Take2 AI founders’ journey, shared in a guest lecture at Stanford Graduate School of Business

When I was a student at Stanford Graduate School of Business (GSB), every startup story I read seemed to be about an overnight success.

Recently, I went back to the GSB to give a guest lecture to MBA students who want to start companies of their own. I wanted them to hear the real version.

The real version is less about a brilliant idea and more about listening and adapting. Kaushik Narasimhan and I met at the GSB, and we built Take2 into the category leader in agentic AI for healthcare recruiting by looking for signals, acting on the right ones, and going all in once we found product-market fit: the point at which founders gain clarity on the solution customers truly need and can’t do without.

Here’s the story I told, and the lessons I shared.

Three Healthcare CHROs, One Question

Take2 started with research. Kaushik spent two years at Stanford conducting IRB-sponsored research on AI and labor market data, and that work became the starting point for our company.

The defining moment came early on, when Kaushik and I were invited to speak on a General Atlantic panel about voice AI agents and the technology we had built. Afterward, three healthcare CHROs came up to us with the same question: could our technology interview their applicants? Their teams were receiving more applications than they could ever handle.

One CHRO asking that question would have been a conversation. Three asking the same question couldn’t be a coincidence. It was a signal, and it defined Take2. Today, we build solely for healthcare, because healthcare HR leaders asked us to build for them.

Why Healthcare Needed AI Recruiting Agents

What those CHROs described is the impossible position healthcare recruiters are in. Hospitals are always hiring, and every opening sets off the same chain of work: sourcing candidates, reaching out to them, screening them, scheduling interviews, verifying licenses, gathering references, and completing paperwork before day one. Meanwhile, applications keep coming in faster than any team can keep up with.

The stakes are high. Clinicians have options, so a slow process means losing them to faster-moving employers. Some roles sit open for months, especially in rural areas, and health systems end up paying external staffing firms to fill them.

Much of that work is logistical and repeats for every candidate, which is exactly what AI agents can now take on. They hold natural phone conversations with candidates 24/7, work inside recruiters’ applicant tracking systems, and move candidates to the next step. That frees recruiters to focus on the conversations that truly need them, those with qualified applicants they want to hire.

Healthcare also adds a layer of complexity that other industries simply don’t have. Prasad Setty, a Stanford faculty member and a Take2 mentor, put it this way:

 “When you’re trying to hire a clinician, there are just so many things that you need to know in terms of the licensing and the credentials that they have. And that’s not easy for a general-purpose AI model to do.”

That’s why every Take2 agent is trained on healthcare hiring data, and every product decision is made with healthcare hiring in mind. It’s a depth that generic solutions built on general-purpose models aren’t designed to provide.

Building with Customers, for Customers

We attribute Take2’s growth to two things: how much pain the product relieves for healthcare recruiting teams, and our customer obsession. That obsession starts with how we build.

We turned our early customers into design partners with a simple promise: “We’re going to build with you, for you.” One of those partners, a talent acquisition leader with 20 years in healthcare who has implemented Take2 at multiple healthcare organizations, met with our team almost every week to review results and talk through what he needed next. That’s a big reason customers choose Take2 today: the product was shaped by people with more than 20 years of healthcare recruiting experience, not by our assumptions about what customers would want.

We still work this way. When a request comes up across several customers, we validate it with others and build it once the need is clear. Many of our six AI agents started that way.

The partnership continues after the contract is signed. Every customer works with a Take2 Engagement Manager and a Forward Deployed Engineer, who set up the agents around the customer’s workflows and optimize them as results come in. Because our engineers handle all integration and setup, connecting the agents directly to the customer’s applicant tracking system, customers onboard in just a few weeks with minimal resources on their end. They aren’t just buying software; they’re buying a partner who will help them iterate until the agents deliver outsized impact.

Where Take2 Is Today

Take2 now serves many of the largest healthcare organizations in the country, including Veterans Affairs, CommonSpirit Health, UCHealth, and Temple Health. Our six AI agents cover healthcare recruiting from first contact to day one:

• AI Sourcer finds candidates for hard-to-fill roles.

• AI Interviewer interviews applicants by voice 24/7 and scores each conversation against a predefined rubric tailored to that specific role.

• AI Scheduler books qualified applicants onto recruiters’ calendars.

• AI Verifier checks licenses and credentials.

• AI Reference Checker gathers references.

• AI Onboarder manages new-hire paperwork from offer to day one.

Take2 customers see up to 70% of recruiter time saved, 2x faster time-to-hire, a 4.8/5 candidate rating, and cost savings on external staffing agencies and sourcing.

What I Told the Students

I closed the lecture with four lessons from our journey. They’re the same principles we run Take2 on today, and they shape how we work with every health system and healthcare organization we serve.

Listen to signals, not praise. People will almost always tell you they like your idea. Pay attention instead to what customers actually spend their time and money on. For us, that signal was three CHROs asking the same question.

Until you find product-market fit, finding it is your only job. And once you find it, go all in. For Take2, that meant focusing entirely on healthcare recruiting. That focus is why our agents understand clinical hiring as deeply as they do, and why leading healthcare organizations choose us over competitors.

Build with your customers, not just for them. Turn early adopters into design partners, and validate every request with other customers before you build it. That’s how we made sure what we shipped worked universally across health systems.

Be obsessed with your customers. Our customers get more than software; they get a partner who helps them deploy AI and fine-tune it until it creates material value.

The best proof of those principles comes from customers themselves. A national recruitment leader at Veterans Affairs texted me this:

“We work with many third-party vendors in the advancement of VA's mission, and you and the team have been the most hardworking partners in my recollection. We appreciate everything, and thank you for your contribution to veteran access to care.”

We see that sentiment across many of our customers, and it’s something we take a lot of pride in.

What’s Next?

For Take2, the journey is far from over. Our customers continue to expand how they use our AI agents, and the needs they bring to us continue to shape what we build. We’re also deepening our relationship with Stanford GSB and are developing case studies that will be taught at the business school in 2027.

As for the students, and any founder still searching for product-market fit, here’s what I hope they remember: it rarely arrives overnight. You find it by listening closely, testing what you hear, and acting with conviction when the signal is clear. That discipline brought Take2 here, and it will guide what comes next.