
Disarray helps ML researchers and engineers automate repetitive ML tasks and accelerate model development. It generates code and automates tasks by using GenAI while integrating with existing data stacks. The platform supports debugging with data, code, and model lineage and enables natural language queries of models, features, and pipelines. Built by experts from UC Berkeley's RISELab, Google, Microsoft, NASA, and LinkedIn, the tool targets enterprise and research teams. Disarray aims to scale ML application development for complex data pipelines and AI models.

Disarray helps ML researchers and engineers automate repetitive ML tasks and accelerate model development. It generates code and automates tasks by using GenAI while integrating with existing data stacks. The platform supports debugging with data, code, and model lineage and enables natural language queries of models, features, and pipelines. Built by experts from UC Berkeley's RISELab, Google, Microsoft, NASA, and LinkedIn, the tool targets enterprise and research teams. Disarray aims to scale ML application development for complex data pipelines and AI models.