Factory and Databricks partner to tie data migration to software modernization

A migration that shaped the partnership

According to Factory, the partnership grew out of joint work for one of the largest global enterprises. The two companies are helping that customer migrate a complex data platform that powers hundreds of business applications to the Databricks Data + AI Platform. The job involves petabytes of data and hundreds of thousands of tables.

The applications include user interfaces, application services, integration and API layers, business logic and rules, security, and management features. Factory says all of these carry dependencies and expectations that must survive the migration. When Factory joined the effort, much of the data had already moved to the Databricks lakehouse. The remaining challenge was rebuilding the applications and workflows that use it.

Factory describes earlier migrations of this scale as resource intensive and risky. Data engineers migrated tables and rebuilt pipelines, while application developers analyzed and wrote new software by hand. The teams used different skills, tools and processes, and even minor changes to a data pipeline created downstream impacts and handoffs.

One engineering workflow, from analysis to deployment

Factory and Databricks say they bring the migration into a single engineering workflow, using agents and automation to understand the existing system, plan the migration, build, and validate the result. Databricks provides the open, governed foundation for data, analytics and AI, while Factory contributes agentic software engineering.

The first step is understanding the current system. Factory Droids work across application code, data pipelines, SQL, tests and available documentation. With access to the relevant repositories and tools, they can trace how an application depends on a stored procedure, which pipeline creates its inputs, and where a change could affect downstream behavior. Factory's context integrations and access to assets governed through Databricks Unity Catalog help connect application dependencies with the data foundation.

That understanding becomes the migration plan. Droids can identify dependencies, propose an implementation sequence and define compatibility checks. Factory's Missions provide planning, milestones and coordinated execution for work that spans multiple components. Engineers review the plan and resolve ambiguous business rules before implementation proceeds.

Once the plan is set, Droids use Databricks API access to create and run jobs, execute queries to validate results, and deploy applications. Factory gives an example: moving an application's operational database from legacy Postgres to Databricks Lakebase, rewriting data transformations in Python, building and scheduling a Lakeflow pipeline, creating an AI/BI dashboard, and comparing the new application and its data against the legacy one.

Three categories of use

The partnership covers three categories of use cases:

• PDLC Automation: Factory can help teams move from a product requirement through planning, development, testing, security, documenting, monitoring and enhancing applications built on or powered by data in Databricks.
• Migration and Modernization: Factory can help teams analyze, plan and execute data platform migrations to Databricks, along with the applications that depend on that data.
• Cybersecurity: Factory can help teams remediate vulnerabilities in applications monitored by Databricks Lakewatch.

Availability and getting started

Qualifying customers can purchase Factory through the Databricks Marketplace using up to 10% of their existing Databricks commitment, according to Factory. The company says it expects many enterprises to benefit from a tighter loop between data and the software it powers.

To begin, users connect Databricks in Factory under Settings → Connectors, then run a task such as building an automation that checks the Lakehouse-to-Lakebase Sync workflow every night and posts a progress recap to the team's Slack channel. Factory's setup guide covers managed MCP servers, running Droids on Databricks-served models, and webhook automations. A demo can be booked on the Factory site.

For enterprises already on Databricks, the offer points to a way to handle application rewrites alongside data moves, with the purchase route through existing Databricks spending. The source does not give pricing or performance figures beyond the 10% commitment option.