Skip to main content
Back to timeline
mastra.aiSource publication:

Mastra launches Custom Factory program: embedded engineers build a software factory around your codebase

Synopsis

Mastra has launched a Custom Factory program in which forward-deployed engineers embed with a customer's team and co-build a software factory for that codebase on Mastra, running in the customer's own stack. Most engineering teams still adopt AI one tool at a time, such as a hand-run coding agent, an issue investigation agent and a code review agent. Mastra says its own software factory now handles 70% of its issues and that its PRs have grown 6x since launch, with most work starting in Slack through Shipyard, its Slack agent, and running through to a pull request ready for review.

Source-provided article image: Introducing Custom Factory Program
mastra.ai

Interpretation

Mastra is offering its own software factory as a starting point, tailored to a customer's organization and processes rather than leaving teams to assemble AI tools one by one. Most engineering teams are still adopting AI one tool at a time: a coding agent someone runs by hand, an issue investigation agent, a code review agent. Mastra says its factory now handles 70% of its issues and that its PRs have grown 6x since launch, a result from running the model inside its own organization.

The factory carries work from issue to a pull request ready for review: it triages and scopes the task, writes the code, runs the tests, checks its own output and brings in an engineer where judgment matters. Each step in the pipeline is run by an agent, and it gets better every time it runs; most work starts in Slack with Shipyard, the Slack agent. The offering includes the underlying factory, a Slack agent for operating it, forward-deployed engineers who embed in the customer's repo and team chat, built-in evals and observability, and human checkpoints with security and governance such as auth and role-based access control.

Perspective

The announcement is aimed at R&D teams evaluating AI engineering pipelines and describes how Mastra's own factory runs plus what the Custom Factory program delivers; the program starts with a whiteboarding session, then co-embeds with the customer team and co-builds until it ships PRs in production.

The 70% issue-handling figure and the 6x PR growth are Mastra's own account of its factory, with no comparison to other teams or prior processes; the material gives no data on whether customer-side factories reach similar results.