Skip to main content
Blog

Engineering

The systems we build, decisions we make, and what we've learned.

Engineering

How a Small Model Learned to Do a Large Model's Job at Docusign's Scale

Running a frontier model on every document doesn't scale at 1M+ contracts a day. In part two of our series on how Docusign contracts agreements at scale, we cover how we trained a smaller, faster model to match — and in production, beat — the accuracy of the larger model it replaced, at a fraction of the cost.

Author Docusign AI Team
Docusign AI Team
How a Small Model Learned to Do a Large Model's Job at Docusign's Scale

Latest posts

Engineering

Turning a Loose Field Definition Into a Reliable Contract Extraction

Docusign processes more than a million documents a day, but 50+ standard extraction fields can’t cover everything an enterprise needs to pull from its contracts. In part three of our series on how Docusign processes contracts at scale, we cover how customers define their own extractions and run them on the same pipeline as our standard models.

Author Docusign AI Team
Docusign AI Team