Meet Preplyze – Your New No-Code Data Processing & Visualization Powerhouse

Preplyze is a cross-platform visual data workflow application built for speed, safety, and simplicity. Users can visually build scalable data pipelines, perform transformations, and create visualizations — all with an intuitive no-code interface, backed by a Rust-powered backend and a React-based frontend. AI-powered features enhance productivity by recommending queries, generating charts, and even creating full workflows.

Core Capabilities

• Visual Workflow Editor

  Drag-and-drop interface to build data workflows using nodes like input, filter, transform, join, etc.

• AI Assistant (via Model Context Protocol)

  Natural language queries → SQL, visual pipelines, chart suggestions, explanations.

  • Generate pipelines from plain English

  • Write or optimize SQL queries

  • Suggest transformations

  • Recommend the right chart type for your data

• Data Processing Engine (in Rust)

  Native Rust performance using libraries like:

  • Polars for dataframe operations

  • Arrow and Parquet for columnar data exchange

  • Optional multithreading or async jobs

• Visualization Engine

  React + Plotly.js, VegaLite, or Recharts on the frontend, optionally generated via LLM.

• Cross-Platform Desktop App

  Built using Tauri — a secure, lightweight Rust shell around the React frontend with a built-in Rust backend API.

Architecture Overview

Frontend

• Framework: React + TypeScript

• Libraries:

  • React Flow (for visual DAG/workflows)

  • Shadcn/UI or Chakra UI (component styling)

  • Plotly.js / Recharts / VegaLite (charts & graphs)

• Platform: Tauri (bundled as native app for Mac, Windows, Linux)

Backend (Rust)

• Core Language: Rust

• Framework: Axum or Actix-Web (for API and WebSocket support)

• Data Engine:

  • Polars (high-performance dataframe library in Rust)

  • Arrow / Parquet (intermediate binary formats)

  • DuckDB / SQLite or similar local SQL engine (optional, via FFI)

• Workflow Engine:

  • Rust-based DAG executor, managing pipeline steps as tasks/nodes

  • Multithreaded for concurrency

AI Layer (Model Context Protocol / Assistant)

• MCP Server: Can be embedded in Rust or run separately (e.g. Python or Node-based assistant server)

• Functionality:

  • Generate workflows

  • Auto-suggest transformations

  • Generate SQL from prompts

  • Recommend visualizations

• Model Access:

  • OpenAI Assistants API, Claude, or self-hosted LLM

  • Uses tool-calling architecture for integrating functions

Tech Stack

Communication

Data & Storage

Planned Directory Structure

preplyze/
├── frontend/                 # React UI
│   └── components/, views/, utils/
├── backend/                  # Rust backend (Axum/Actix)
│   ├── engine/               # DAG processor, executor
│   ├── dataframe/            # Polars handlers
│   ├── mcp/                  # Optional AI assistant integration
│   └── routes/               # API endpoints
├── assets/                   # Icons, logos
├── workflows/                # User DAGs & pipelines
├── data/                     # Cached outputs (Arrow, Parquet)
├── storage/                  # SQLite metadata DB
└── src-tauri/                # Tauri config and Rust entry

Who Is Preplyze For?

• Data Engineers who want reusable, inspectable pipelines

• Data Analysts who prefer drag-and-drop over code

• Scientists and Researchers who need fast prototyping

• Educators & Learners building mental models of data pipelines

Value Proposition

Conclusion

Preplyze is a blazing-fast, AI-powered visual data platform that runs locally. Built on Rust and React, it offers an intuitive no-code interface backed by high-performance engines — making complex data processing accessible, safe, and fast for everyone.