KartikChugh/Otto

Leverages Wit.ai for natural language task inference and model recommendation, parsing user objectives into ML categories (regression, classification, NLP) and automatically selecting appropriate algorithms based on fuzzy keyword matching against a curated model database. Features an interactive neural network designer, real-time data visualizations, and generates annotated Python code throughout the pipeline—enabling users to learn by example while building models through a conversational interface rather than traditional code or configuration forms.

961 stars. No commits in the last 6 months.