[ Announcements ]
TextQL Now Supports kdb+

TextQL Now Supports kdb+

Ivy Cho, Benjamin Shi, Joseph Ma

Jun 15, 2026

TextQL Now Supports kdb+

kdb+ is the database that holds market data at Goldman Sachs, Morgan Stanley, JP Morgan, and much of the rest of Wall Street. But getting an answer out of it requires writing q, and the people who write fluent q are few, expensive, and perpetually backlogged. Today that changes. TextQL connects to kdb+, and Ana speaks q so your team doesn't have to.

Removing the quant bottleneck

Pulling anything out of kdb+ used to require a quant that’s fluent in q. Now, the same data answers to anyone who can type a question.

Ask in plain English. Ana writes the q, runs it against your kdb+ instance, and hands back a chart, a report, or a number the same way she works across the rest of your stack.

One more connector matters less than who can use the data behind it: a portfolio manager, a risk lead, a compliance analyst, anyone who'd otherwise wait on a quant to translate their question. And because Ana already joins across your stack, a question spanning tick data in kdb+, positions in Snowflake, and counterparty records in Postgres becomes one query instead of three exports and a spreadsheet.

Built to handle q honestly

Connecting properly is harder than pointing a generic SQL adapter at kdb+, and the difference shows up in whether you can trust the answer:

  • q's own operations run end to end. Take the as-of join, which pairs each trade with the quote that prevailed at the moment it executed. Treat it like SQL and it gets corrupted. Ana lets q do what q does, so it renders correctly down to the last detail.
  • Missing data stays missing. In q, an absent value is a sentinel rather than a blank. A naive reader turns a missing price into a very large or negative number, poisoning every aggregate above it. Ana translates these back into true nulls.
  • Hardened by default. kdb+ ships no driver for our setup, so we wrote our own client that reads untrusted bytes behind strict safety checks.

See it on your data

Add the connector under Connectors → Databases → kdb+. If your firm runs it and that data has always lived a layer away from the people who need it, this is the bridge.