MLStudio
A guided, web-based machine learning workspace for projects, visual workflows, experiments and model serving.
MLStudio is a Flask web application for creating machine learning projects, uploading or importing Kaggle datasets, building node-based workflows, and tracking and comparing experiments. It ships 80 reusable workflow templates in 17 categories, from generic pipelines to oil and gas, healthcare, finance and retail. Workflows run on cron schedules or file triggers and can call OpenAI, Anthropic, Ollama, Groq, Mistral or Azure OpenAI models. The companion MLServer serves each model in its own environment, with Prometheus metrics, drift monitoring, promotion and audited access control.
Features
Projects with experiment tracking, side-by-side comparison, snapshots and workflow versions
Dataset upload, preview and profiling, plus Kaggle browsing and download
Visual workflow builder with scikit-learn, evaluation, SQL, HTTP, S3, LLM and vector nodes
80 workflow templates in 17 industry and generic categories
Cron scheduling and file-watch triggers
LLM nodes for OpenAI, Anthropic, Ollama, Groq, Mistral and Azure OpenAI
MLServer: isolated per-model environments, Prometheus metrics, drift monitoring, promotion, access control and audit
Publishing of projects and datasets to AI Community, and entry into its competitions