Contact me
Case studies Reference system

SaaS MVP with AI at its core

From a validated idea to a live product: onboarding, subscription, user area and the AI feature that justifies the price.

For
Founders and startups
Service
Build + AI
Hover over or tap a block to see its role in the system.

Reference system: a need I come across often, documented with the exact architecture I would put in place. Other client work is presented in meetings.

01Problem

The idea has been validated with early customers, but there is no CTO and no tech team.

Every month without a product is a month of lost traction, and hiring a full team before the first revenue is a risky bet.

02Context

Users
First B2B customers and the founding team.
Current process
Manual demos, a spreadsheet behind the scenes, scattered tools.
Constraints
A limited budget, a useful V1 quickly and then iterations, clean code that a future team can take over.
Tools in place
  • Mock-ups
  • Spreadsheet
  • Isolated no-code tools

03Solution

A V1 focused on the core promise: onboarding, user area, the AI feature that justifies the subscription, and recurring payment.

Everything else waits for real usage signals.

Why this approach. An MVP proves one thing. Anything that does not serve that proof slows down the launch.

04Architecture

Interface

  • Web app Landing page, onboarding and user area, deployed on Vercel from GitHub.

Data

  • Auth Sign-up, sign-in, teams: Supabase Auth.
  • Supabase Product data, usage quotas and subscription status.

Automation

  • Functions Server logic: AI calls, quota checks, webhooks.
  • n8n Onboarding emails, activation reminders, internal notifications.

AI

  • Claude The feature at the heart of the value, with structured output and quality control.

Output

  • Payment Recurring subscription through a payment provider's API.

05Features

  • Sign-up and guided onboardingUsers reach the value in their very first session.
  • User areaTheir data, their results, their account.
  • Core AI featureThe reason to pay, with reliable, verifiable outputs.
  • Subscription and billingRecurring payment, free trial, plan changes.
  • Automated emailsWelcome, activation, reminders: the product guides users on its own.
  • Minimal back officeJust what is needed to follow accounts and usage.

06Target result

  • First paying users on the real product.
  • Clean, versioned code that a future team can take over.
  • A short feedback loop.

How it is measured

  • Activation of new accounts
  • 4-week retention
  • Trial-to-paid conversion

07What I do

  1. Scoping the V1
  2. UX/UI
  3. Frontend and backend
  4. Database and authentication
  5. Payment and AI integrations
  6. CI/CD and production launch
Next case study Custom operations back office

A system like this one, for your business?

Contact me All case studies