AI / SAAS PLATFORM
SparkUp AI
An AI workflow automation marketplace of ready-made tools, with a drag-and-drop builder for assembling new ones without code.
- CLIENT
- SparkUp AI
- INDUSTRY
- AI / SaaS Platform
- STACK
- React · TypeScript · .NET · Azure · MUI · React Flow · Stripe · SurveyJS · EditorJS
OVERVIEW
The project
SparkUp AI is a marketplace for AI workflow automation, a catalogue of ready-made tools called Sparks that a business applies to a routine task without building anything itself. A user states a goal in plain language and gets a result back, with no prompt engineering in between. Allzone worked on the platform that hosts, configures and runs those tools, and on the builder that lets people assemble new ones.
CHALLENGE
What made it difficult
A marketplace of automation tools has to solve two problems at once: give each tool enough configuration to be genuinely useful, and keep the experience simple enough that a non-technical user can apply one in minutes.
Behind that, the platform needs a consistent execution model, so that tools built at different times, doing different things, all run through the same infrastructure with the same handling of inputs, outputs and failures.
WHAT WE DID
The work delivered.
- 01
Marketplace platform
Catalogue, discovery and configuration surfaces for the tools available on the platform.
- 02
Workflow execution
A consistent model for running automation tools, handling their inputs and returning structured results.
- 03
Dynamic form and content handling
Configuration and content interfaces built with SurveyJS and EditorJS to support varied tool inputs without bespoke UI per tool.
- 04
Spark Builder
A drag-and-drop canvas, built on React Flow, that compiles a visual workflow into an executable automation, so an advanced Spark can be assembled without writing code.
- 05
Creator marketplace
Listing, sharing and payments, so the people building Sparks can monetise them and the catalogue grows from both sides rather than from one fixed toolset.
- 06
Application engineering
A React and TypeScript frontend against a .NET service layer, deployed on Azure.
ARCHITECTURE
How it fits together
- React + TypeScript application layer with a Material UI component system
- .NET service layer handling tool definition, execution and results
- Each Spark independently deployable, so the marketplace grows without every tool sharing one codebase
- More than one model provider behind a common orchestration layer, so a Spark runs on whichever model suits its task
- Structured configuration via SurveyJS; rich content authoring via EditorJS
- Azure hosting and platform services
TECHNOLOGY
The stack
- React
- TypeScript
- .NET
- Azure
- MUI
- React Flow
- Stripe
- SurveyJS
- EditorJS
OUTCOME
What exists now.
A new automation tool reaches the catalogue without a bespoke frontend build, because configuration adapts to the tool rather than the other way round
A non-technical user applies a tool to their workflow in minutes, without engineering support
Someone with no engineering background assembles an advanced automation on a canvas, then lists it for other people to use
The platform keeps adding tools without the execution model underneath them needing to change
I must say that Allzone's development team was fantastic to work with! They took the time to fully understand our needs, handled every detail with care, and communicated clearly throughout the project. Everything was delivered on time, and the final result exceeded our expectations. Truly a reliable and talented team to partner with.
NEXT
Related work.
NEXT STEP
Tell us what you're building.
Bring us the problem with its real constraints attached. We will tell you what we would build, and what we would not.