"Most of your data in an organization is actually in the form of this unstructured data - 90% plus of most organizations - it's data in an unstructured form, but it contains an awful lot of very useful structured data as well," said Kus."
"Generative AI was born on unstructured data, so it can understand these things, and this is where you suddenly have the ability to not only use and analyze unstructured data in its raw form, but also the next generation of capabilities to extract structured data," Kus said."
""We've been looking at this for a long time and while you used to be able to get labels out of image or train a model to ... do handwriting recognition and OCR. But it was very limited on what it could work on, because you had to very specifically train these very specific models," he continued. "Now, with something like Box Extract, we can pull almost any structured information out of almost any file type ... almost like a human would do.""
Box introduced three AI capabilities: Box Automate, Box Extract, and Box Shield. Box Extract uses generative AI to locate and pull structured information from diverse unstructured file types. Box Automate deploys an AI agent to streamline multi-step business processes and reduce manual specialist checks. Box Shield enhances data security while enabling greater value extraction from organizational data. The capabilities target the large proportion of enterprise data that exists unstructured and aim to convert that data into actionable, structured information to improve analysis, workflows, and control without bespoke model training.
Read at IT Pro
Unable to calculate read time
Collection
[
|
...
]