The CloudWinks framework

Build software from definitions, not from scratch.

CloudWinks is an application framework for business software. An application is described in data — its screens, its fields, its menus, its tables — and the framework builds the working product from that description. Every application it builds shares the same shell, the same identity layer and the same pipeline.

Definitions-driven · A shared .NET service API · PostgreSQL data layer · Flutter client on web and phone

What it is

One framework underneath every application.

Most software grows into a collection of separate products that each need their own login, their own interface and their own maintenance. CloudWinks was built the other way around: a single framework holds the shared parts, and each application contributes only the definitions that make it different.

An application registry, not a bundle

The framework is the container. It registers the applications, holds their state, and serves their navigation. Applications are entries in that registry, which means one deployment can carry many products on one database server.

Metadata-driven interface

Every screen, field and menu is a row in the database, not a hand-coded page. The client asks the framework what to draw and renders it. Change the definition and the interface changes with it — for every user, everywhere.

One schema per application

Each application owns its schema inside a single PostgreSQL server, so data stays cleanly separated while operations, backups and reporting stay in one place. A shared identity base ties users and companies across every application.

Sign in once

Authentication is part of the framework, not an add-on. Sign-in goes through Firebase, with providers for Google, Facebook, LinkedIn and X configured at the framework level. Google is the one in daily use on the development environment today.

Modules that are already modelled

The platform applications carry the hard, unglamorous work: identity, files and media, notifications, reporting, messaging, integrations, workflow and scheduling. A new product starts from those definitions rather than from a blank page.

One pipeline, repeatable

A Python build, import and seed pipeline materialises an application from its definitions: schema first, then screens, then data. Rebuilding a clean environment is a command, not a project.

How it works

From a description to a running application

An application is specified as numbered definition folders and an application_setup.json that declares its name, its category and the hostnames it answers on. The pipeline does the rest, in a fixed order.

Build

Creates the schema, tables, functions, views and constraints for the application from its table definitions.

Import

Registers the application, then imports its controls, fields and menus so the framework knows how to draw it.

Seed

Loads the starting data, including the rows that must be tied to the application's own identifier.

Serve

The shared client opens the application, in a browser and on a phone, behind the one sign-in the framework provides.

The shape of the stack

Three languages, each doing the job it is good at: Dart and Flutter for the client that runs in a browser and on Android, a shared .NET service API as the gateway to the data, and Python for the build pipeline. PostgreSQL is the single source of truth.

  • A Flutter shell renders every application, on web and on mobile, from the same definitions.
  • A generic API endpoint behind one gateway means applications do not each invent their own backend.
  • Per-application schemas share one PostgreSQL server, one backup policy, one operations story.
  • Firebase provides authentication and file storage, so identity and uploads work the same everywhere.

What that buys you

The expensive part of business software is rarely the first screen. It is the twentieth screen, the fifth customer's permission model, and the third year of maintenance. A framework that treats every screen as data changes that economics.

  • A new industry application is a set of definitions, not a new project.
  • Fixes and improvements reach every application they apply to at once.
  • One visual language across products, so users learn once.
  • A clean environment can be rebuilt from the definitions at any time.

The platform layer

The shared services nobody has to build again

These are the shared services of the ecosystem — the parts every serious business product needs, modelled once so no application writes them twice.

CloudWinks

The framework itself: the registry, the shell, the navigation and the shared interface components.

Crux

The identity and company base that every other application extends for its own users and organisations.

KeyKook

Authentication and access: credentials, sessions, roles and multi-factor sign-in, applied consistently wherever a person can reach.

Murphy

Error logging, grouping and audit, with a client-side queue so an error survives a dropped connection. Built into the platform, not yet switched on in a running application.

MediaMogul

Files, images and media, stored through the shared storage providers and served to any application.

TheReporter

Reporting and output, so applications produce documents and exports rather than reinventing them.

BigTalk & BullChat

Messaging and conversation inside an application, using the same identity as everything else.

WorkerBee & TheBoss

Background work and orchestration: jobs that run without a screen in front of them.

Roady

Maps and geospatial work: routes, pins, polygons and geofences, with location tracking for people who work away from a desk.

TheIntegrator

Connections to outside systems, with mappings verified as part of the build.

GoldenRule

Shared business rules, kept in one place so applications behave consistently.

CoderCody & Cloudify

Development and deployment tooling that lives inside the platform rather than beside it.

See the full list of platform and industry application definitions →

Industry applications

Application definitions across ten sectors

Each one is modelled to the way that industry actually works, and each inherits the platform underneath it. Maturity varies widely between them, so read the catalogue as what the framework can describe, not as a shelf of finished products.

Agriculture

  • FarmDance
  • FoalFactory
  • HorseStudbook
  • HuntingLodge
  • StockMan

Financial

  • BeanCounter
  • DreamDrafter
  • GrabAtunity
  • LandMan
  • PutAway
  • TimeTracker
  • TjingTjing
  • Zibit

Business specific

  • BeautyBlast
  • FathersFlock
  • Pharmakologie
  • ShootingClubs
  • StableKidz
  • YumYum

Management

  • DealMaker
  • Equipico
  • ProcureFlow
  • SellWell

Project & task

  • GoalGetter
  • RapidResolve
  • TaskMaster

Home & family

  • ComfyCottage
  • Familia
  • GoGrownup
  • SmartKidz

HR

  • BizzFrame
  • SkillScraper

Hobbies & sports

  • FireSticks
  • HorseShowBizz
  • MuscleMadness

Web platforms

  • SellToday
  • Surfy
  • UseAbit

Government

  • MrSam

Definitions, not code

A table, a field, a screen and a menu are records in the database. Adding one does not mean writing or rebuilding a feature by hand.

One pipeline

Build, import and seed run in a fixed order from one orchestrator, so a clean environment can be produced again from the same definitions.

One identity base

Applications extend a shared record of users and companies, which is why the same account works across everything the framework carries.

Where it stands

In active development, and honest about it

CloudWinks is a working development system, not a finished product. The framework, the database model, the build pipeline and the client shell are built and running in an active development environment, and application definitions exist that build and seed against it.

What that means in practice: if you want a product built on CloudWinks, you are getting a foundation that already exists, plus the work of turning the right definitions into the product you need. That is a very different conversation from starting at an empty repository — and a different one from buying finished software off a shelf.

We would rather tell you exactly where the edges are at the start than discover them together in month six.

Built and running

  • The framework shell, navigation and registration stack.
  • The database model: each application in its own schema, on one server.
  • The build, import and seed pipeline, rebuilt clean from scratch.
  • The generic query API and authenticated gateway.
  • Web and Android builds of the client from one codebase.

Still being finished

  • Depth of individual applications beyond their modelled screens — maturity varies widely between them.
  • Production hosting and multi-tenant isolation.
  • Third-party integrations beyond those already mapped, and error monitoring switched on in a running application.

Tell us what your business actually does.

We will tell you which parts of it CloudWinks already covers, which parts are a definition away, and which parts are genuinely new work.

Start that conversation