Ship better
models faster

AI-generated code is great, but how do you make it work with “path/to/your/data”?The answer is not RAG or MCP.

Meet Disarray, an AI pair programmer that truly understands your data, brought to you by ML researchers and engineers working on the cutting edge to reimagine ML development.

Generate code that works with your data

Join select ML engineers rethinking their workflows with agentic tooling.

Automate the repetitive

Skip the tedious tasks with code generated to work with your data stack, so you can focus on modeling.

Debug with advanced lineage

Trace issues across your entire data stack and see how data, code, and models connect.

Answers from your stack

Get instant, cited answers about your models, features, and pipelines with natural language.

Reuse what works

Surface relevant code, models, and experiments – build on collective knowledge and avoid duplication.

Behind the vision

We built foundational tools for AI's first wave and are now channeling over a decade of battle-tested expertise to reimagine ML application development with GenAI.

Our founders, Doris and Moustafa, met at UC Berkeley's RISELab, the innovation hub behind Databricks. Our team brings together ML researchers and engineers from Google, Microsoft, NASA, and LinkedIn.

We're proud to have the support of renowned advisors and investors, including the founders of Databricks, Kaggle, CDOs, and the former US Chief Data Scientist. We're backed by top institutional investors.

Doris Xin

Moustafa AbdelBaky

Daria Mehra

Rodrigo Curbelo

Jeff Boehm

Andy Konwinski

Heidi Lanford

Joel Young

Anthony Goldbloom

DJ Patil