YC published their Spring 2026 Request for Startups and I read it the way I read every RFS: with one eye on what they're signaling and the other on whether I'm already building any of it.
The overlap with healthcare fintech is larger than I expected. AI for product management, government AI, stablecoins, AI developer tools. All of these touch problems I'm either working on or watching from close range. Let me walk through the ones that caught my attention.
AI for product management. YC wants AI that replaces the product manager's workflow: user research synthesis, prioritization, specification writing. I didn't apply to YC for this, but I accidentally built a version of it for dental practices. Dentplicity compresses thousands of market signals into a handful of actionable moves. The practice owner is the PM. The AI is doing research synthesis, competitive analysis, and prioritization. The difference is that our "product" is the practice itself, and the "roadmap" is a set of marketing and operational decisions. If someone built this for software teams with the same specificity we bring to dental markets (entity-resolved, geo-indexed, time-series), it would be very good. Most PM tools are horizontal. The vertical version, built for a specific type of decision-maker with specific data inputs, is where the real value lives.
Government AI. The RFS mentions AI for government services, permitting, procurement. I've spent enough time reading OCC memos, CFPB enforcement actions, and BSA/AML guidelines to know that the compliance side of healthcare banking is basically a government AI problem. Regulatory interpretation, document generation, audit preparation. All of it could benefit from AI that understands the specific regulatory corpus. What I'd build if I weren't building Dentplicity: an AI compliance officer that ingests OCC, CFPB, and state banking regulator publications and generates gap analyses against your current policies and procedures. The CCO I described in My Biggest Non-Technical Risk costs $250K. An AI-assisted version might cost $50K plus the tool. That's a meaningful change in the economics of compliance for pre-seed fintech founders.
Stablecoins. YC is interested in stablecoin infrastructure for payments and financial services. Cross River is already expanding stablecoin ramps. The dental banking intersection is narrow but real: cross-border payments for lab work (a lot of dental lab manufacturing happens overseas), stablecoin-denominated B2B payments that settle faster than ACH, and programmable escrow for insurance receivables. I'm not building this. But if the stablecoin rails mature to the point where a dental practice can pay a lab in Shenzhen without a wire transfer fee, someone should integrate that into a vertical banking product.
AI developer tools. I have strong opinions about this one because I live it. The tools that helped me build Dentplicity (Claude Code, Cursor) changed the economics of bootstrapping fundamentally. I wrote about the specifics in Bootstrapping in the Age of AI Agents. What's missing from the current generation: context persistence across sessions, understanding of business logic beyond code structure, and the ability to operate within constraints I've set rather than requiring me to review every output. The chat-first interface problem I described in Chat-First Interfaces Are a Dead End applies to developer tools too. I want to set a direction and have the tool execute, not have a conversation about every file.
What I'd build if I weren't already building. Honestly, I'd build vertical AI compliance for regulated industries. The problem is massive (compliance infrastructure eats 10-15% of operating budgets), the data is structured (regulatory publications, enforcement actions, exam procedures), and the buyers are motivated (nobody wants to spend $250K on a CCO if an AI-assisted version achieves the same exam readiness). Healthcare, banking, insurance, legal. Each vertical has its own regulatory corpus and its own set of compliance obligations. The company that builds the vertical compliance AI for one regulated industry and then expands to others has a real platform play.
But I'm building Dentplicity and CLIN. The RFS is interesting to read from the sidelines. It confirms that the problems I'm working on (AI-powered decision intelligence for specific verticals, financial infrastructure for healthcare) are in the water. Whether that's validation or just pattern-matching is a question I'll answer by shipping, not by analyzing RFS documents.