What stack will my project be built on? And why the chosen one isn’t about impressing any reviewer.
Why the right startup-stack question isn’t "what stack" but "what shape do we ship on deliverable day?" — and how a written scope determines the stack before any invoice is sent.
The right question is “what shape on deliverable day?”
The first time a Series A founder hears the word stackused as a quality signal, the right answer is to ignore it. A stack is a means, not an outcome. What the brief is really asking is: what does the finished, shipped, deployed surface look like on deliverable day — and does it match the brief? The chosen stack is the answer to thatquestion, not the answer to a code-reviewer’s taste.
That shift — from acoustic-guitars-on-the-resume to shipped, measured, billed result — is the whole point of the agency model. The stack is chosen to ship, not to impress any reviewer on a portfolio call.
TypeScript, Next.js, Prisma, Postgres, Stripe Billing.
The stack under this very website is the same one the agents ship on for clients: TypeScript on a Next.js front-end, a Prisma layer against Postgres, Stripe Billing for the checkout primitive, and a small set of opinionated frameworks that keep the day count honest. The case studies on the site document the variations — TypeScript + Next.js + Prisma + Postgres for a D2C rebuild, TypeScript + Node + Algolia + Postgres + Redis for a search-relevance fix — but the spine of the stack stays the same.
Reading the stack list top-to-bottom is the wrong exercise; reading it in terms of “what shipped, what was measurable, what was the pricing tier” is the right one. The same principle runs through the four questions on the FAQ, where the stack, the IP question, and the proposal-round question are answered in the same register.
The agents work on the founder’s stack, not a new one.
The stack question resolves differently when the brief is on rails. The agents don’t ship a fresh stack to the founder — they work inside the founder’s existing repo, with the founder’s branch policy, on the founder’s provider accounts. Work-for-hire from day one: every commit in your repo, every deploy on your account, every PR mergeable by you with a single click. The stack is the founder’s; the artefacts are the founder’s; the IP is the founder’s. We don’t hold a proprietary model trained on your work, and we don’t keep a private corpus of your code after the engagement ends.
See how the pair of choices — outcomes, not hours; your repo, your IP — shapes the pricing page on the pricing page. The tier shape sets the cadence, the response SLAs, and the published ship-date estimate, all in writing.
The five questions on /scope decide the stack before any invoice is sent.
The five short questions on /scope walk through the parts of the brief that decide the stack before any invoice is sent. Current repo or greenfield, the data shape, the auth model, the deploy target, and the brief’s tier are the only five inputs; the rest of the scope is the answer to those five. A written scope lands back within 72 hours, and it’s revised — not invoiced — if anything in it is wrong.
The full intake brief lives at /intake; the tier-by-tier pricing cadence lives on the pricing page. Send the rough idea; the rest writes itself.
Liked the read? Send the work, get a scope back.
Send the rough idea through the intake form. We come back with a written scope within 72 hours, and we don’t send invoices until you’ve signed off on it.
Or email wingman-stack@polsia.app with a one-paragraph brief.