Khroma is an innovative AI-powered color discovery tool designed to help creators and designers find, refine, and organize the perfect color schemes for their projects. Developed by George Hastings, the platform leverages machine learning to provide a highly personalized experience, ensuring that the color palettes generated align specifically with the user's aesthetic preferences. By training a neural network directly within the browser, the application learns individual color tastes to produce unique and relevant combinations.

The functionality centers on a unique training process where users select fifty preferred colors to calibrate a neural network. This algorithm, which has been pre-trained on thousands of popular human-made palettes found across the internet, then generates an infinite variety of color combinations. These outputs can be viewed in several formats, including typography layouts, gradients, flat palettes, or applied to custom images, allowing users to visualize how colors interact in real-world design contexts.

Some of the key features are:

• AI Personalization: Uses a neural network trained by the user to generate custom-tailored color palettes that exclude disliked colors.
• Infinite Generation: Creates an endless supply of color combinations based on a massive database of human-curated design palettes.
• Advanced Search and Filtering: Allows users to filter combinations by hue, color type, tint, value, hex codes, or RGB values.
• Detailed Color Metadata: Provides comprehensive information for every palette, including color names, hex codes, RGB values, CSS implementation code, and WCAG accessibility ratings.
• Accessibility Awareness: Includes a specific feature to block color combinations that fail to meet WCAG contrast ratio requirements for readable text.
• Collection Management: Supports the creation of an unlimited library of saved color combinations for future reference and project use.
• Data Portability: Enables users to export and import their personal generator data, allowing for consistent use across different devices and browsers.

The operation of Khroma is straightforward, beginning with the user choosing fifty colors to build their initial preference model. Once the training is complete, the generator displays various color pairings that the user can explore and refine. Users can interact with the generated results through a search bar that supports natural language-style queries, such as searching by specific hues like "muted yellow" or exact hex codes. Settings allow for further customization, such as toggling chromatic gradients or enforcing strict accessibility compliance to ensure designs remain functional and legible. The entire process is designed to be fast, intuitive, and highly responsive to user input.

Some common use cases include:

• UI/UX Design: Quickly finding accessible and aesthetically pleasing color pairs for web interfaces, buttons, and backgrounds.
• Branding and Identity: Developing consistent color palettes that align with specific brand moods, whether they are neon-bright or deeply muted.
• Graphic Design: Generating inspiration for posters, illustrations, and print materials by viewing colors applied to diverse typography and gradient presets.
• Accessibility Auditing: Ensuring that foreground and background color combinations meet legal accessibility standards for web readability.