Services · SaaS
Full SaaS builds for founders who ship.
A SaaS product is not a landing page with a Stripe button. It is auth, billing, a dashboard people return to, integrations, deliverable email, observability, and infrastructure you can hand to a future engineering team. We build the whole thing.
What a SaaS build actually covers
Auth: email plus a single OAuth provider, password reset flow, magic links if the audience prefers it, session and refresh token handling, account deletion to satisfy GDPR and the App Store reviewer when there is a mobile companion. Billing: Stripe subscriptions, metered or seat based, dunning flows, invoice history, tax through Stripe Tax when the customer base is global. Dashboards: not a single chart bolted on, but the pages that a paying customer opens daily. Integrations: usually two or three at launch, picked from the customer interviews.
Plus the parts that do not appear on a feature list: rate limiting, observability through OpenTelemetry into Sentry or Honeycomb, structured logs, error budgets, a deploy pipeline that lets the future team ship Friday at 4pm without flinching.
Our default stack
Next.js 16 App Router with React 19, TypeScript strict, Tailwind v4, server components by default with client islands where interaction requires them. Supabase for auth, Postgres, storage, and realtime. Drizzle ORM for typed queries. Stripe for billing. Resend or Postmark for transactional email. Vercel for the web layer with edge functions for low latency routes. Inngest or Trigger.dev for background jobs and scheduled work. Sentry for errors, PostHog for product analytics.
When the workload outgrows Supabase, we migrate to managed Postgres on Neon or Crunchy, replace Supabase auth with WorkOS or a hand rolled auth service, and keep the same Drizzle layer so application code does not move. We do not pick exotic stacks for the novelty.
When to hire a studio vs build in house
Hire a studio when you are pre Series A, have fewer than three engineers, and the time to first paying customer matters more than building internal headcount. Studios trade a flat fee for compressed time. A SaaS that would take a small in house team nine months can ship in twelve to sixteen weeks with a studio because we are not also hiring, onboarding, and writing engineering ladders.
Build in house when you already have a senior engineering team, a stable architectural point of view, and the runway to absorb the slower start. The handoff from a studio to an internal team is part of our engagement. We write the README that an in house team will actually read.
What we do not build
Crypto. Trading platforms. Anything where the value prop is regulatory arbitrage. Apps that exist to game search rankings. We do build commerce SaaS that sits next to a Shopify stack, AI native products that need a Fractional AI Officer, and B2B SaaS that needs a Fractional CTO for the next two years.
Process and price shape
Two week discovery: customer interviews on file, a written architecture doc, a deploy ready repo with CI, a Linear backlog with the first six weeks scoped. Then ten to fourteen weeks of build with a weekly written update and a shipped increment every Friday. Final two weeks: handover, runbook, oncall coverage if the team wants it for the first month, and a written exit doc.
One slot per project. Cash, equity, or a blend. The intake closes at the end of Q3 2026. Six slots across the studio.
FAQ
Common questions
What does a SaaS engagement usually include?+
Auth, billing through Stripe, a working dashboard, the core feature set, at least two third party integrations, transactional email, observability, and infrastructure provisioned in your accounts. We hand over a system that does not need us to keep running.
Why Next.js plus Supabase as a default?+
Next.js gives us server components, edge runtime, and a single language across the stack. Supabase gives us Postgres, row level security, auth, storage, and realtime without us hand rolling six services. When a project outgrows Supabase we migrate to managed Postgres and split auth and storage out cleanly.
When should I hire a studio vs build in house?+
Hire a studio when you have less than three engineers, a clear product wedge, and a six to twelve month window to get to the first paying cohort. Build in house when you already have a senior engineering team, an active product, and the time to hire and onboard.
How do you handle multi tenancy?+
Schema per tenant for regulated and high isolation cases. Shared schema with a tenant id column and Postgres row level security for standard B2B SaaS. We pick on the first call based on data sensitivity, scale, and customer expectations.
Do you ship MVP or production grade?+
Production grade. We do not use MVP as an excuse to skip auth hardening, billing edge cases, or rate limiting. The first version is small in scope, not loose in quality.
What about AI features in the SaaS?+
When AI is core to the product, we pair the SaaS engagement with our Fractional AI Officer scope. Eval pipelines, model routing, and cost control are folded into the build.
Related
Keep reading
Next step
Have a SaaS to ship?
Tell us the wedge, the customer, and the timeline. We reply from a real inbox within two business days.