Although the API gives you all the building blocks to find and exploit your data, the flexibility of such approach is limited as opposed to a domain specific language (DSL).
In addition, an API is not always convenient when you need to compose requests, perform joins or work interactively on the database.
The goal of the quasardb querying language is to provide a syntax close to the SQL language so that users become quickly productive while leveraging the unique features of quasardb typically absent from relational databases such as native timeseries support and tags. All of this while abstracting away all the inherent complexity of distributed computing.
Queries are meant to extract small subsets of data (thousands of entries) out of very large data sets (billions of entries), fast enough to be used interactively (in a couple of seconds).