What is a data lake?

A data lake is a catch-all repository for raw data. Where a warehouse insists on clean, structured tables, a lake accepts almost anything — database extracts, CSVs, JSON logs, PDFs, images — and stores it in its native format so nothing is thrown away. That flexibility is the appeal, but it's also the weakness: a lake that no one has organised becomes a "data swamp" that's hard to actually get answers out of, because structure and meaning still have to be applied later before you can query it.

iDBQuery overlaps with the lake idea but approaches it from the question end rather than the storage end:

• It handles mixed data. iDBQuery connects to structured databases and warehouses, spreadsheets and CSVs, REST APIs, and even whole folders of documents and PDFs — reading scanned files with OCR.
• It applies meaning automatically. A semantic layer teaches it what cryptic tables and columns mean, so you don't have to hand-curate the raw pile before asking anything.
• It answers in plain language, with citations. Rather than leaving you to write queries against a sprawling store, iDBQuery writes the SQL, runs it against one live model of your sources, and cites every figure back to its exact source.

So if a lake is about keeping everything, iDBQuery is about getting cited answers across everything — including the messy document and file data that lakes are often used to hold.