Document Intelligence for Regulated Pipelines
High level talk on how to use artificial intelligence to do document classification and named entity recognition on regulated pipeline documents. We use NLP and NER to classify 120 different pipeline documents spanning many formats and extract common meta data from them so the documents can be stored and easily retrieved when needed.
Pairing this advanced technology with an experienced scanning team can digitize a room full of paper faster than other approaches. We walk through this approach over a more common rules based approach using things like Regex or pattern matching and why using more advanced techniques like NLP is a superior approach.
