AuroraML launches FrameForge: real-time model editing for production pipelines

AI · 5 min read

AuroraML today announced FrameForge, a visual model editing platform designed to replace manual code changes across production ML pipelines. The tool provides a live view of model graphs, data flows, and deployment endpoints, enabling engineers to make and test structural edits without touching container orchestration configs. FrameForge supports model versioning, canary rollouts, and integrates with major MLOps platforms.

Alongside the product launch, AuroraML disclosed a $45 million Series B led by Sequoia with participation from existing backers and new strategic investors from the edge computing space. The company said funds will go toward expanding enterprise support, hiring across product and sales, and accelerating integrations with cloud providers and inference accelerators.

Early customers report reduced mean time to patch for production models by as much as 60 percent, saying the visual interface helps cross-functional teams collaborate during incident response. AuroraML plans to offer a free tier for research teams and a hosted enterprise plan with single-tenant deployment options for regulated industries.