SciStudio

SciStudio is a desktop application for scientific workflows. It brings data, scripts, apps, and AI assistance onto a single visual canvas where you can design pipelines using typed blocks and watch them execute. The goal is to consolidate the scattered tooling that data scientists and researchers deal with daily: notebooks here, scripts there, custom utilities somewhere else, results in a folder you can't find.

The core interface is a workflow canvas. You drop in blocks, connect them, and the system handles execution order and data flow. Blocks are typed, so you can see what goes in and what comes out. That makes it easier to catch mismatches before you run a pipeline rather than debugging halfway through. Python is bundled, so you don't need to manage environments separately.

AI assistance is built in at two levels. There's a conversational interface that helps you build and debug workflows. You can describe what you want, and it suggests blocks and connections. Then there are AI agent blocks, which are pipeline components that can inspect data structure, make decisions, and produce reproducible outputs. The AI layer is powered by Claude Code or Codex depending on configuration.

The project awareness feature is worth understanding. SciStudio maintains context across your files, notebooks, scripts, and custom tools. That means the AI assistant knows what you're working on rather than treating each request as isolated. For complex analysis projects with many moving parts, that context helps the assistant give relevant suggestions instead of generic ones.

It's aimed at researchers and data scientists who spend too much time switching between tools. If your workflow involves Jupyter notebooks plus shell scripts plus a custom Python module plus some downloaded utility, this is an attempt to put all of that in one place where you can see the whole pipeline and run it reproducibly.

SciStudio is open source under the MIT license. You can download it from GitHub Releases or clone the repo and build it yourself. It runs on macOS and Windows. Community support is available through GitHub, Slack, and Discord. There's no paid tier because the whole thing is free.