[](https://devin.ai/blog)

# Devin for Databricks Migrations: Less time migrating, more time building

CognitionNovember 19, 20263 min read

![Devin for Databricks Migrations, with Devin and Databricks logos](https://devin.ai/images/blog/devin-for-databricks-migrations/cover.jpg)

Today we’re launching [Devin for Databricks Migrations](https://devin.ai/partnerships/databricks), built with Databricks to move enterprises off legacy data platforms and onto the Databricks Data + AI Platform. Migrations finish faster so engineering teams get their time back for higher-value work. Once their data is on Databricks, teams have a unified, open, and governed foundation for analytics, AI applications, and agents.

### What Devin does on a Databricks migration

Devin, the AI software engineer, is now being used by enterprises to move legacy data platforms to Databricks. On a Databricks migration, Devin converts code from the legacy data warehouse, tests it in the customer’s workspace, fixes what fails, and reconciles every source against its Databricks copy before cutover. Devin can autonomously run hundreds of tasks in parallel, which is how one Fortune 500 energy company migrated a 142-object platform to Databricks Unity Catalog in just eight days. With data governed through Unity Catalog, enterprises can more quickly make their data available for analytics and AI workloads.

> “AI is only as good as the context you give it. For some companies, critical business knowledge is stuck in legacy systems. With Devin and Unity Gateway, customers can easily move context from legacy platforms into Databricks, with everything governed in Unity Catalog, so they can start building AI that actually understands their business,” said Ali Ghodsi, Co-founder and CEO of Databricks.

SourceDevinDatabricks

Teradata

Oracle

SQL Server

Informatica

Redshift

Snowflake

Databricks SQL

PySpark

Spark Declarative Pipelines

Unity Catalog

Reconciled before cutover

- Schema
- Row counts
- Aggregates
- Row-by-row diff

### A four-week project in two and a half with Zeb

A global digital transformation strategy company, Zeb, used Devin to move NetSuite data into Databricks, finishing a four-week migration in two and a half weeks.

> “The question we ask on every migration isn’t how many engineers we can put on it. It’s what the highest-leverage team looks like for the outcome the customer needs. For a global investment advisory firm, Devin and our engineers built the NetSuite-to-Databricks pipelines and 20 tables, and we delivered a four-week project in two and a half,” said Mal Vivek, CEO at Zeb. “We’ve committed our Databricks delivery teams to Devin because it gives our customers more value per dollar.”

### Why now

Migrations have been a major barrier between enterprises and the AI applications and agents they want to build. Devin for Databricks Migrations takes on work that used to tie up a whole engineering pod for multiple quarters and is able to parallelize tasks to accelerate project execution.

> “We built Devin to own software engineering projects end to end. Migrations are a great example where agents need to understand the existing code, make the changes, and test that everything works,” said Scott Wu, Co-founder and CEO, Cognition. “This typically takes a lot of engineering time. With Devin for Databricks Migrations, we’re helping teams get through it faster so they can start building with their data sooner.”

We partner with Databricks as a strategic inference provider, running large-scale open source and frontier model workloads with the reliability our customers expect. Because Databricks provides access to a broad and growing range of models through a single endpoint with Unity Gateway, we can move quickly as new options emerge, choosing the right model for each need and adapting to shifts in demand without re-architecting our stack.

### Get started

Devin for Databricks Migrations is available today on the [Databricks Marketplace](https://marketplace.databricks.com/details/9c82aa78-ac29-4a36-a179-4ad16bf72b93/Cognition_Devin-by-Cognition), and customers can apply their purchase toward their existing Databricks Universal Commit. Devin is now one of Databricks’ preferred migration tools, and customers can work through their account team to run migrations with their own engineers, with Databricks field engineers or with delivery partners such as Zeb. Through Databricks Marketplace, customers can access partner solutions that help accelerate their data and AI initiatives while maximizing the value of their existing Databricks investments.

Learn more at [devin.ai/partnerships/databricks](https://devin.ai/partnerships/databricks).
