← All work
Healthcare · United Kingdom · LLM / Text-to-SQL

Medicine shortage intelligence

Natural-language querying over live medicine shortage data — staff ask in plain English and get a real answer, no SQL required.

847/day
NL→SQL queries

Around 847 natural-language queries a day, answered directly from live data — analyst-grade access for the whole team.

Context

Frontline pharmacy staff needed to interrogate fast-moving shortage and tariff data, but only a handful of people could write the queries to do it.

Challenge

Make a complex, frequently-changing dataset answerable by anyone on the team, accurately, without exposing them to SQL or stale reports.

What we built

  • A text-to-SQL layer translating plain-English questions into validated queries over the live dataset.
  • Retrieval and guardrails so answers stay grounded in the actual data, not the model's guesses.
  • A clean query interface surfacing results, not raw tables.

Outcome

Around 847 natural-language queries a day, answered directly from live data — analyst-grade access for the whole team.

Stack

LLMText-to-SQLPythonPostgreSQL

Ready to ship

Let's get your production AI live.

Bring us your operational data. We'll build the system that acts on it — deployed in your environment, measured in production.