We read every YC Requests for Startups batch, extracted the highest-signal ideas, and turned each one into a complete founder playbook — with customer profiles, 8-week MVP plans, pricing models, and 90-day action plans ready to execute.
EVERY IDEA INCLUDES
Specific, financially quantified, described from the customer's perspective on a specific bad day. No vague pain points.
The exact timing argument: what changed in 2025–2026 that makes this buildable today but wasn't possible 2 years ago.
Exact job title, company size, budget authority, and where to find them. Not "SMBs" — the specific person who signs the check.
Week by week: what to build first, what to validate, what to skip entirely. Ship something real in two months.
Pricing tiers, unit economics, revenue framing, and annual contract value estimates grounded in real market data.
Specific channels, specific messages, specific communities, and outreach scripts that actually get replies.
Who exists, what they charge, exactly where they fall short — and the precise wedge where a new entrant wins.
The specific things most likely to kill this business in the first 12 months, with concrete mitigation approaches.
Day by day, from zero to first paying customer. Actual tasks, sequenced, with decision gates.
+ 4 BONUS SECTIONS INCLUDED WITH EVERY PURCHASE
How to validate any idea in 2 weeks without writing a single line of code
How to price your AI startup correctly from day one
The fastest path to your first $10,000 in revenue
The complete technical stack to build any AI product in 2026
YC SUMMER 2026 RFS
01 · AGRITECH
AI vision and precision robotics to identify and treat individual weeds and pests in real time. Biology replacements — microbes, peptides, RNA — replacing entire classes of synthetic chemicals. Farmers stuck in a bad loop. This breaks it.
Written by Garry Tan
02 · SCIENCE
Beyond copilots. Intelligent systems that run closed discovery loops: propose hypotheses, design experiments, analyze results, iterate — in drug discovery, materials science, and protein engineering.
Written by Jon Xu
03 · SERVICES
Don't sell software — sell the outcome. Insurance brokerage, accounting, tax, audit, compliance, healthcare administration. The services market is many times larger than SaaS.
Written by Gustaf Alströmer
04 · HEALTHCARE
Agent harnesses analyzing genome scans, EHRs, wearables, and diagnostics for user-specific treatment. Sequencing costs falling faster than Moore's Law. FDA more open to n-of-1 therapies than ever.
Written by Ankit Gupta
05 · ENTERPRISE AI
The missing layer between raw company data and reliable AI automation. Pulls knowledge from Slack, tickets, email — structures it into an executable skills file that every AI agent can use safely.
Written by Tom Blomfield
06 · DEFENSE
A Patriot missile costs $3M. An FPV drone costs $500. High-capacity interceptors, sensor fusion OS, non-kinetic defenses, attacks on the autonomy stack itself. Drone defense is becoming a distributed systems problem.
Written by Tyler Bosmeny
07 · SOFTWARE
Coding agents let users become their own forward-deployed engineers. Software companies ship shared primitives; users radically customize the final interface for their exact workflow.
Written by Ankit Gupta
08 · SPACE
Reusable rockets creating massive demand for space compute. Inference chips optimized for mass, thermal efficiency, and radiation tolerance. SpaceX and NVIDIA experience directly relevant.
Written by Philip Johnston
09 · HARDWARE
Shenzhen: design to physical part in a day. US: weeks. Startups that produce parts dramatically faster and tightly integrate design, manufacturing, and logistics into one fast loop.
Written by Nicolas Dessaigne
10 · SPACE
Extracting silicon, aluminum, iron, and titanium via electrolysis on the moon. 3D printing from molten regolith — more efficient than on Earth because no support structures needed in low gravity.
Written by Adi Oltean
11 · CHIPS
Agents loop: calling tools, branching, backtracking, holding context. Current GPUs hit 30–40% utilization. Purpose-built silicon for the agent execution loop — and the compiler that makes the chip work.
Written by Diana Hu
12 · SAAS
AI collapsed software production cost by 10–100x. The legacy moat is gone. Clone and price at 1/10th, rebuild AI-native from scratch, or bundle ten point solutions into one suite. Go after chip design, ERPs, industrial control.
Written by Jared Friedman
13 · INFRASTRUCTURE
The next trillion internet users are AI agents. They need machine-readable interfaces: APIs, MCPs, CLIs — not forms and buttons. Every major software category needs a rebuild with agents as first-class citizens.
Written by Aaron Epstein
14 · ENTERPRISE
F100 buyers are actively seeking AI solutions for the first time ever. YC companies landing multimillion-dollar deals during their batch. A 2–3 person team can ship something a Fortune 10 finds useful before the ink is dry.
Written by Harshita Arora & Brad Flora
15 · CHIPS
1,400 process steps, a dozen countries, 5 months — managed with spreadsheets. Real-time allocation tracking, multi-tier risk monitoring, export compliance. CHIPS Act fabs building supply chains from scratch.
Written by Diana Hu
TIMING
YC Summer 2026
YC's Summer 2026 theme is rebuilding software, services, and silicon from scratch and pushing AI into the physical world. The most consequential batch for founders in a decade.
The window is open
AI cut software production cost by 10–100x. Moats that took decades to build are now vulnerable. The window to enter and win is open — and it won't stay open long.
Save 100+ hours
Reading every RFS, mapping the thesis, identifying the customer, building the MVP plan — done. You get straight to the only question that matters: is this the right idea for me?
GET THE PLAYBOOK
One purchase. Every future batch update included. Start reading in the next 2 minutes.
One-time · PDF download · Lifetime updates as new RFS batches drop
Not affiliated with Y Combinator · Ideas sourced from publicly available YC RFS pages