Scaling Document Processing Across the Harvey Platform

Harvey's document-processing redesign separates work that fails for different reasons. Its engineering account describes distinct stages for extracting document contents, turning text into searchable representations and writing those representations into a retrieval system.

Each stage can retry and scale independently. A slow scanned document can occupy extraction capacity without blocking indexing, while a database rate limit can pause indexing without forcing extraction to repeat. Durable workflow state also lets interrupted work resume, and individual bad files can fail while the remaining batch continues. The architectural benefit is containing delays and failures as document collections grow.

“The main lesson was simple: At this scale, document processing stops being one pipeline and starts behaving like several different systems.”

“The latest complete week in the measurement window averaged about 3.5 million documents per day. Compared with a year ago, that is 26x more documents and 39x more data.”

“Instead of relying on one-off uploads, users are increasingly treating us as a central system of record for massive document collections.”

“Growth was not just more documents; it was more kinds of documents, with more OCR, conversion, and indexing edge cases showing up at production frequency.”