Call Prep: Kevin Zhang (Bain Capital Ventures)

TL;DR

Kevin's thesis is agentic software that reimagines services with machine intelligence. His portfolio company Pallet is the exact analogy — domain-specific AI agents deployed across complex operational workflows. DaisyAI is the healthcare version of that story. He already believes in the model. The call is about: are we the right founders to execute it in healthcare? Lead with the Pallet parallel, lean into the relationship, and show domain depth.

His Profile in 30 Seconds

Partner at BCV for 10 years (since 2016)
Thesis: agentic software, reimagining services with machine intelligence
Board seat at Pallet (agentic AI for logistics — closest comp to DaisyAI)
Led seed in Tandem Health (healthcare)
Previously first business hire & Head of Product at Fundera (→ NerdWallet)
Columbia BA, co-founded Columbia In Tech
Style: long-term relationships → invest. Knew Pallet founder 5 years before leading seed.
Thomas connected ~1yr ago. Michael knows him personally.

Investor Type: Thesis-Driven Generalist with Healthcare Exposure

Per our segmentation, Kevin is a healthcare-aware generalist whose first question will be about scalability and the tech-vs-services tension. But because his thesis is literally "agentic software replacing services," the usual friction point (are you a consulting firm?) is actually his core belief.

Likely Questions & Best Answers

1. "How is this different from what Pallet does, but for healthcare?" (HIGH probability)

Best answer: "Same playbook — domain-specific AI agents deployed into complex operational workflows. Pallet does logistics, we do clinical operations. UM review is the wedge: highest volume, most painful, most structured. Our AI conducts the clinical analysis, the nurse reviews and decides. We land with one workflow, prove ROI, expand across the payer's AI roadmap."

Why he'll ask: He'll naturally pattern-match to his portfolio. Make it easy for him.

2. "What's reusable vs. custom for each client?" (HIGH probability — this is THE question)

Best answer: Tau framework. "Three revenue components: SaaS platform fee (5-10%), implementation on a rate card at a discount to Palantir, and ongoing maintenance/licensing (~50-60% of contract). The core analytical layer — criteria matching, evidence extraction, recommendation engine — is reusable. Implementation is payer-specific. Each deployment gets faster."

Why he'll ask: Scalability is the core tension in his thesis. Pallet solved it (70+ customers). He wants to see our path.

3. "Tell me about Premera — what does the engagement look like?" (HIGH probability)

Best answer: Wharton connection → embedded as their AI engineering team → started with UM review → on their 12-initiative AI roadmap. FDE model: we're inside building, not outside selling. This is how Pallet works too — embedded in the customer's operations.

Why he'll ask: He wants to understand the deployment model. The FDE/embedded framing will resonate given Pallet.

4. "How big is this market?" (MEDIUM probability)

Best answer: This is a gap in our prep. Have at least a rough framing: "X thousand UR nurses nationally, $Y billion in payer admin costs, AI restructures all of it. UM is the wedge, healthcare admin back-office is the platform."

Why he'll ask: BCV is a large fund. He needs to see this can be a big outcome.

5. "What's your conviction — why UM specifically?" (MEDIUM probability)

Best answer: "We spent a year in the weeds — conferences, nurse interviews, payer conversations. UM is the highest-volume clinical review workflow, it's deeply structured (criteria-based), the financial stakes are massive (every review is a pay/deny decision), and the talent crisis is acute. It's the perfect wedge for agentic AI in healthcare."

Why he'll ask: "I appreciate founders with independent conviction." Show him the depth of your thesis.

6. "Why now? What's changed?" (MEDIUM probability)

Best answer: Three shifts converging: (1) AI models are finally good enough for clinical text analysis, (2) payer cost pressure is accelerating post-COVID, (3) nursing shortage means they can't hire their way out. Every investor we've talked to validates this timing from their own networks.

Why he'll ask: Standard question, but he'll appreciate that we've validated it across multiple investor conversations.

7. "How do you think about defensibility?" (LOW probability — but be ready)

Best answer: Three layers: (1) Healthcare domain depth — knowing WHERE to apply AI in clinical workflows is the hard part, (2) compounding data — every review makes the system smarter for that payer's patterns, (3) switching cost — once embedded in their data systems and team workflows, ripping us out is harder than keeping us. Same as Pallet — the moat is being embedded, not the model.

What to Lead With

Name the Pallet parallel immediately. Don't make him figure it out — say it:

"You already have the thesis for what we're building. Pallet is agentic AI for logistics operations. We're agentic AI for clinical operations. Same playbook: domain-specific agents deployed into complex workflows where the status quo is manual labor."
Then go into UM as the wedge, Premera as proof, and the expansion path.

This skips past the "what do you do" phase and puts you in a conversation about execution and domain depth — which is where you want to be.

What to Avoid

Don't pitch "healthcare AI" generically. Kevin explicitly avoids hype. Don't sound like every other health-AI company.
Don't undersell the relationship. Kevin invests in people he knows. The fact that Thomas has been in touch for a year and Michael knows him personally is a feature, not a footnote. Be warm, be real.
Don't over-explain the tech. He was Head of Product at Fundera — he's technical enough. Focus on the domain insight and the business, not the architecture.

Demo Strategy

Probably not needed for the first call. Kevin is thesis-driven and business-model focused (like Sam at Tau, like Chris at Sorensen). Save the demo for a follow-up if he wants to go deeper on product. If he asks to see it, have it ready — but don't lead with it.

The Relationship Play

Kevin's investing pattern is: know founders for a long time → invest when the timing is right. You're at the "timing is right" moment:

Thomas connected ~1 year ago (relationship established)
Michael knows him personally (trust layer)
The thesis match is obvious (Pallet = logistics, DaisyAI = healthcare)
You have a real client (Premera) — it's not vaporware

This call should feel like catching up with someone who already gets it, not a cold pitch. Be conversational. Let the relationship carry some of the weight.

Answer Gaps to Watch For

From our synthesis:

Market size / TAM — biggest gap. Kevin's at a large fund. Have at least rough numbers. This is where we've been weakest.
Competitive landscape — he may not know the UR/payer AI space. Be ready to frame it clearly.
Written materials — if he read the deck, watch for the "wrapper" impression. Use the Tau reframe if needed.

Last updated: Feb 12, 2026