Datasaur is an artificial intelligence company specializing in creating solutions for data labeling. The company builds a data labeling workforce management platform for Natural Language Processing (NLP), a branch of artificial intelligence that helps computers understand, interpret, and manipulate human language. The technology can be used for business purposes in cases concerning misinformation detection, contract summarization, product review analysis, customer service call transcriptions, and receipt and invoice understanding. Datasaur is based in Sunnyvale, California and was founded in 2019 by Ivan Lee.
Datasaur’s software has six basic functions: named-entity recognition, coreference resolution, tagging parts of speech, dependency resolution, document labeling, and image classification. Named-entity recognition allows for the identification and extraction of key entities for automated content discovery. Coreference resolution is the process of finding all expressions in a text that refer to the same entity. Tagging parts of speech allows for the software to label each individual word with its part of speech. Dependency resolution is the process of breaking down the grammatical structure of a sentence. Document labeling allows for the categorization and classification of data. Lastly, image classification allows for the comparison, classification, and annotation of images, videos, and more. The major use cases for Datasaur’s software are product review analysis, receipt and invoice understanding, customer service call transcription, contract summarization and understanding, misinformation detection, and direct message and forum moderation.
Datasaur snags $3.9M investment to build intelligent machine learning labeling platform
September 29, 2020
Datasaur, a semi-automated text data-labeling tool, raises $1 million
March 2, 2020
Launch HN: Datasaur (YC W20) - data labeling interface for NLP | Hacker News