These days, companies first drill into their data set before they drill for gas or gold. We know that the analysis in the first drilling step deals with huge volumes of data today, but data variety is also a growing challenge.

JSON is a data format that can represent a variety of data types, so it is quickly becoming the lingua franca of data in the Internet for APIs, data exchange, data storage, and data processing. In the business intelligence world, however, SQL is the language to analyze the data. Hence, there are many “SQL-on-Hadoop” projects.

Traditional SQL isn’t friendly to non-relational data. Apache Drill unifies SQL with JSON, CSV, Parquet, and other complex/multi-structured data types in Hadoop . It enables interactive analytics using your favorite BI and analytics tools directly on your data without ETL/upfront modeling requirements. In this talk, we’ll introduce Apache Drill as the technology to drive the audacious goal of bringing self-service data exploration natively on Hadoop, and discuss its use cases.