Database management systems rely on an implicit pact: they provide quick and correct answers in exchange for precise, complete and syntactically correct questions. In most cases, database users express those questions using a query language such as SQL. However, this model does not work as well in cases of data exploration, because users cannot specify exactly what they want and where to get it. This is because they have little or no knowledge of the data involved, and their requirements are often vague and abstract. Thibault Sellam asks how such users can match the precision required by a query language. He presents four assistive systems that can help users compose and refine interesting new queries: Claude, Blaeu, Ziggy en Diamond.