Data Navigator Uses NLP for Queries
A new AI-based data “navigation” system is billed as the first analytics platform to use natural language processing to generate SQL queries from spoken requests. The result is a sort of virtual assistant for data analysts.
Promethium, a data management startup based in Menlo Park, Calif., said this week its NLP-generated SQL queries can be used to assemble data across different systems and platforms. The navigation system is based on the company’s AI and machine learning-driven contextual automation tool. The software can be used to locate data, then present it in form that is most useful to business analysts.
Founded in February 2018, Promethium said users can ask a question, prompting its system to locate data and demonstrate how it should be assembled. The system then generates an SQL statement to access the desired data before executing the query.
The federated queries are designed to run across all databases, data lakes and warehouses with sub-second query times. The system also identifies duplicate data and data provenance to help ensure compliance with governance rules.
The vendor said one customer deployed its NLP-based data navigator to automate manufacturing processes and reduce its data team to a single analyst. The resulting data query times across multiple systems were reduced from month to minutes, Promethium claims.
“It’s like having a virtual assistant that allows anyone to put the value of their data to work,” added Kaycee Lai, Promethium’s CEO and founder.
The self-service analytics platform is available as a software service via public cloud or in a customer’s private cloud. An on-premise version also is available.