It's mostly backend code in Java. Almost no UI although they are adding a bit now. Their product takes unstructured data like text documents and extracts meaning out of it. It is not NLP but more mathematical algorithms and models that establish relevance of terms and documents to each other. The work will include quite a bit of R&D and understanding/learning of various algorithms related to content analysis and parsing and categorization. It is not a typical project where you get some data from the user, store it in the database and search for it later. Technologies used would be core Java, Guice, Hadoop, PostgresSQL, Jetty.