Skip to content

Background

Mipmap3D builds photogrammetry software for surveyors, architects, and construction teams — turning drone imagery into accurate 3D models. They had a functional desktop application, but the distribution model was a dead end: every customer required a full license purchase, installations had to be managed per machine, and there was no way to onboard users at scale.

They needed to move to the web. A cloud-based product would let them offer monthly subscriptions, reach users without an install, and build a real acquisition funnel. We were brought in to design and build the web frontend and backend API that made that transition possible.

The Challenges

  • Desktop-only distribution required a full license purchase per user — no trial, no subscription path
  • No web presence meant no self-serve onboarding; every sale was manual
  • The existing Qt desktop codebase wasn't portable to the browser — a full frontend rebuild was required
  • Processing workloads needed to move server-side without changing the output quality users depended on
  • The team needed to keep the desktop product running while the cloud version was built in parallel

Our Approach

We rebuilt the product surface in two layers: a web frontend users could access from any browser, and a backend API that handled processing jobs server-side.

Phase 01: Web Frontend — Replacing the Desktop Client

We designed and built a React/Next.js web application that replicated the core workflow of the desktop client: project creation, image upload, job submission, and results download. The goal was feature parity with the existing app, delivered entirely in-browser with no installation required.

  • React + Next.js web app replacing the Qt desktop client
  • Project management dashboard with real-time job status
  • File upload and results download — no local installation

Phase 02: Backend API — Moving Processing to the Cloud

We built a backend API layer that accepts job submissions from the web frontend, queues processing tasks, and returns results asynchronously. This decoupled the user-facing product from the compute layer and made it possible to scale processing independently of the frontend.

  • REST API for job submission, status polling, and results retrieval
  • Async job queue for processing tasks
  • Cloud-hosted compute — users pay for what they use, not a per-machine license

Phase 03: Subscription Model & User Accounts

With the product on the web, we wired up user authentication, account management, and a monthly subscription billing flow. Users could now sign up, start a free trial, and upgrade — without any manual sales involvement.

  • User authentication and account management
  • Monthly subscription billing integration
  • Self-serve onboarding — sign up to first job without human intervention

The Results

01

Desktop-only product successfully migrated to a web-based SaaS

02

Monthly subscription model live — replacing per-license sales

03

Self-serve onboarding enabled for the first time

04

Processing moved server-side — no installation required for end users

05

Foundation in place for Mipmap3D to build a scalable acquisition funnel

Final Takeaway

Moving from desktop to cloud isn't just a technical migration — it's a business model change. By rebuilding Mipmap3D's product surface as a web SaaS with a subscription billing flow, we gave them a product they could actually sell at scale.

Technologies We Use

Modern, proven technologies to build robust applications

React

React

Next.js

Next.js

Node.js

Node.js

TypeScript

TypeScript

A

AWS

Docker

Docker

Find your best next step

01 —Your situation