a23one

mlpotion

Changelog

Dataset Editor, Model Training & Prediction Playground Enhancements

Dataset editor tools, async model training, simplified prediction playground

This release adds agent-driven autofill and data editing tools for the dataset editor including row inspection, improves UI/UX for the dataset editor with zebra striping and better user feedback, launches asynchronous background model training to avoid blocking user actions, and simplifies the prediction playground and API endpoints with deferred model loading.

  • added

    Added agent-driven autofill and data editing tools for the dataset editor, including row inspection

  • changed

    Improved the dataset editor UI/UX with zebra striping, better feedback and streamlined training workflows

  • added

    Launched asynchronous background model training, so training runs without blocking user actions

  • changed

    Simplified the prediction playground and API endpoints with deferred model loading

Marketing Landing Page & Project Documentation Updates

New marketing landing page, updated project README and setup guides

This release launches a new animated marketing landing page with a six-stage workflow visualisation and call-to-action buttons, plus updates to the project README with new product details, setup instructions and environment variable information.

  • added

    Added a new marketing landing page with animated hero, six-stage workflow visualisation and call-to-action buttons

  • changed

    Updated the project README with new product details, setup guides and environment variable info

Chat Assistant & API Enhancements

Chat endpoint, chat toggle, health check, prediction API, model download, diagnostics, playground, training presets

Users can converse with an AI assistant, use the prediction API, and receive diagnostics and improved training options.

  • added

    Added async chat endpoint for Strategy Consultant AI agent.

  • added

    Added a persistent bottom-right floating chat widget for the AI assistant with streaming SSE responses and saved session history

  • added

    Added chat session listing and context loading, letting users access and reload past conversation sessions

Core UI & Account Management

Landing page, base template, dataset UI, deployment UI, stylesheet, account pages, settings

Users can sign up, view a welcome page, manage datasets, train models, and deploy them.

  • added

    Added landing page that redirects authenticated users to the dataset list and shows a welcome screen for guests.