Most agencies want a 6-month engagement. Most freelancers want a small slice. I ship the smallest version of your product that proves the thesis — fast — so you can put it in front of real users before you spend the next round. Riya.chat: 7 days. LeadCube: under a week. Garment-OS Phase 1: 1.5-2 months.
Founders pitch the V3 to engineers, who scope the V3, who quote 6 months, and then everyone wonders why the runway burned without a single user. I rip the idea apart in the first conversation — find the one thing that has to be true for the business to exist — and we ship only that. Everything else gets scheduled for Phase 2.
The "MVP" you described is actually a V3. The real MVP is one workflow. Maybe two screens. I'll show you which one.
Most early-stage builds don't need a 5-person team. They need one architect with AI multipliers, shipping every day.
The market will teach you something. A 6-month build can't absorb that learning. A 4-week build can.
Most agencies disappear after delivery. You need someone who'll think about Phase 2 while building Phase 1 — so the architecture absorbs change.
I challenge the idea. Find the one thing that has to be true. Cut everything else.
Cheapest path to a working MVP. Buy-vs-build decisions. Existing tools where they fit (Shopify backend, ERPNext, etc).
Daily progress. Weekly check-ins. Functional product running in your hands.
Onboard first users. Watch behavior. Phase 2 scope informed by data, not guesses.
WhatsApp AI companion. Built solo on Node.js + WhatsApp Business API + OpenAI + Redis state. Founder had the idea and no technical background. We stripped it to one thing: get the AI talking to users on WhatsApp. Everything else came later. Now also live on Telegram.
Self-hosted CRM with cloud calling integration. Built the auto-dialer, inbound call tracking, and self-serve onboarding in under a week. Three enterprise clients live, each saving 3x vs LeadSquared.
Could have built custom infrastructure. Didn't. Used Shopify as the backend — got reliability, payments, and negligible server costs out of the box. Non-technical founder went from manual phone bookings to fully digital. The right MVP often uses tools you didn't think were meant for it.
Raw material management system for a garment manufacturer. Scoped from on-site discovery. Phase 1 ships the critical loop (bill → roll → rack → cutting → defect → vendor return) — everything else is Phase 2. Architecture absorbs the future without paying for it now.
Cutting scope is the most valuable thing I do in week one. You'll thank me when the deadline holds.
Cursor, Claude Code, agentic workflows. I move at small-team velocity solo. Which means you don't need to fund a team to find out if the idea works.
I use Shopify, ERPNext, Twilio, Auth0 — whatever already solves part of the problem. Custom only where custom matters.
Not "we'll productionize later." Auth, audit logs, payments, deployment — done right the first time. The same architecture runs at 7,500 users week one.
Most of my best MVPs were for founders who couldn't write code. I translate, scope, build, and hand over with documentation they can actually use.
Phase 1 ships the thesis. Phase 2 unlocks scale. Phase 3 is multi-tenant or platform. The architecture knows about all three on day one.
20-minute call. Bring the idea, the pitch, the deck — whatever's closest. By the end of the call you'll have a scoped MVP plan and a timeline.