MVP development
An MVP you can launch, and keep.
Acasa Labs builds MVPs for founders: the smallest product a customer will pay for, on web, mobile, or both, with AI where it is the point. We have launched five products of our own, four live and one in early access, so we scope an MVP like people who will have to run it afterwards.
The problem
Most MVPs are either too big to ship or too fragile to keep.
One kind of MVP takes nine months because every idea made it into scope. The other ships in three weeks on shortcuts that have to be thrown away the moment customers arrive. Both waste the runway. A good MVP is small in features and sound in the parts that are expensive to change later: data, accounts, billing, and the API.
What we build
- 01
Scope cut to the first paying customer
We write down who pays, for what, and the one job the product has to do on day one. Everything else goes on a list for after launch.
- 02
Foundations you keep
Accounts, a real database schema, an API the mobile app can use later, and billing that can grow into plans. The parts that hurt to rewrite are done properly once.
- 03
Web, mobile, or both
A Next.js web app, an Expo app for iOS and Android, or both on one backend, depending on where your customers are.
- 04
Launch, measure, iterate
Analytics and error tracking from the first release, so the week after launch is spent on what users did, not on guessing.
From production
What our own MVPs grew into.
One product, five surfaces
NoteTubeNoteTube runs on the web, iOS, Android, a Chrome extension, and a remote MCP server. They all sit on one backend, with one FastAPI service for AI.
A second product on the same base
GymFastGymFast runs gyms, and GymFast Pro gives independent trainers the same tools for their own clients. Both use the same platform and mobile API.
A backend built to carry more
SigmaCPMSigmaCPM's multi-tenant backend, with organizations, seats, and per-module entitlements, is shared with a sibling product and its Chrome extension.
Providers you can swap
KraftrKraftr's generation and storage providers are pluggable by config, so the product can change models or move storage between GCS and Cloudflare R2 without a rebuild.
Honest trade-offs
When you should not build an MVP yet
If you have not spoken to ten potential customers, do that first. If the question is whether anyone wants it, a landing page and a waitlist answer it in a week. If the question is whether you can build it at all, for example a model's accuracy on your data, run a technical spike before a product. We will tell you which one you need.
What we reach for
- Next.js, React, TypeScript
- Expo, React Native
- Supabase, FastAPI, or Django
- Postgres
- Stripe, Razorpay, RevenueCat
- OpenAI, Anthropic, Gemini
- PostHog, Sentry
Proof
Shipped, not pitched.
01Our product
NoteTube
A lecture becomes a study system: chat, flashcards, quizzes, all grounded in the source.
An AI learning platform that turns PDFs, YouTube lectures, recordings, and notes into a structured notebook with grounded chat, spaced-repetition flashcards, and graded quizzes.
Web · iOS · Android · Chrome extension · MCP server
02Our product
GymFast
Members, billing, classes, and workout plans for a gym, in one platform and three apps.
Gym management software for gym owners and personal trainers. A web dashboard and a staff app run members, plans, invoices, enquiries, and classes; members get their own branded app; GymFast Pro gives independent trainers the same tools for their clients.
Web dashboard · iOS · Android · Staff, member, and trainer apps
03Our product
Kraftr
One flat-lay photo becomes catalog-ready 4K on-model imagery.
An AI catalog studio for fashion and jewelry brands. Upload one flat-lay or hanger shot per SKU and get on-model imagery at up to 4K, in ten aspect ratios, sized for Amazon, Myntra, Flipkart, and Shopify.
Web · iOS · Android
04Our product
SigmaCPM
Construction project management where AI reads the tender and shows its work.
One project record for civil contractors, from tender to closeout: bids, contracts, RFIs, vendor quotes, and drawings. The AI pulls line items and bond terms out of tender PDFs and cites the page for every value.
Web · API · Chrome extension
Questions
MVP development questions we get asked.
How long does an MVP take?
It depends on the scope, which is why the first call is about cutting it. We give you a written timeline with the scope before you commit, and we hold to the scope rather than letting it grow silently.
Do you work with non-technical founders?
Yes. You bring the customer and the problem. We handle the product decisions on the technical side and explain each one in plain language.
What happens after the MVP launches?
You can keep us on to build what the first users ask for, or hand it to your own team. Either way, the MVP is built so the second version is a set of changes, not a rewrite.
Can the MVP include AI?
Yes, when AI is the point of the product. NoteTube, Kraftr, and SigmaCPM are all AI-first products. We also tell you when a simpler rule would do the job.
Who owns the code and the accounts?
You do, from day one: the repository, the cloud account, the app store accounts, and the billing account.
MVP development
What is the smallest thing your customer would pay for?
Bring the idea and who it is for. We will help you cut it to a first version and tell you what it would take to ship.