We started out as a consultancy helping data teams at investment firms make better use of Bloomberg data.

While working with Data Analysts we noticed that even those who were expert Python/R programmers commonly used Excel as a secondary tool — particularly for communicating computations to teammates, fast-turnaround prototypes, calculation scratch pads, and when answering "This doesn't look right?" questions about ETL results.

We also noticed that existing Excel tools were surprisingly primitive when serving Data Analysts who come to Excel with SQL in hand and need easy parameterization and fast edit/run feedback loops.

Optimal(xl) is our solution. It is founded on three central ideas:

• By making it fast and easy to parameterize and execute SQL within Excel — and assuming you'll do anything beyond that with other tools — we can build a simple, Excel-based SQL playground that is dazzlingly better than Power Query or VBA for SQL-centric work.
• In-cell SQL — with connection metadata alongside as a cell comment — which generates or updates an Excel table directly below it via shortcut keys or menu command, is the data-acquisition method preferred by many Data Analysts. It is more understandable, observable and debuggable than Power Query, VBA or bespoke VSTO/COM/XLL add-ins. And it doesn't pollute Excel's calculation chain with relatively slow data-access calls like Google Sheets' QUERY(...) and Bloomberg's BDH(...) worksheet functions.
• The right mixture of withered technologies (Excel C API; VBA) and new technologies (F#; Snowflake; Cloudflare Workers) can produce Excel-based software that is far more robust, performant and enjoyable than Excel users have come to expect.