FluxPrompt canvas and multi-model prompt flow
FluxPrompt editor and node layout
FluxPrompt prompt builder
FluxPrompt React Flow graph
All experience

AI, Prompting, AI agents

Enhanced AI

About the product

FluxPrompt is Enhanced AI's web hub for working with multiple LLMs on a shared canvas: prompt cards, side-by-side runs, model switching, integrations, and saved libraries. The core editor is a large React Flow surface with Figma-style live collaboration over WebRTC so multiple people can work the same graph with low-latency presence and interaction. Built with React, TypeScript, and Tailwind CSS, backed by Node.js and Postgres on AWS.

Engagement
Client project (Motomtech)
Duration
2 yr
Team size
5 people
Platforms
Web

Role & Responsibilities

Technical Lead & Senior Full-Stack Developer

Jan 2022 - Dec 2023
  • Led architecture and delivery of the FluxPrompt canvas: multi-model prompting, real-time previews, libraries, and organized prompt workflows.
  • Shipped 50+ AI-oriented React Flow node types and edge patterns (generators, routers, tool calls, and layout that stayed readable at scale) with custom controls and validation.
  • Built Figma-grade collaborative editing on the graph: shared cursors, selection, and canvas updates with WebRTC so teams could co-build prompts without refresh or merge pain.
  • Owned the prompt library and builder UX, performance of large graphs, and a model-agnostic surface that could absorb new providers, integrations, and feature flags from product.

Technologies

ReactReact FlowWebRTCNest.jsTypeScriptMaterial UIReact Hook FormZustandNode.jsPostgreSQL