AI in Data Analytics Data Self-Service in the publishing business
Data Access Barriers
The Problem
Our client, a large publishing house, uses data to make strategic decisions and inform campaigns. Although the publisher has accumulated a treasure trove of data, access to it is often difficult, especially for non-technical staff. The SQL code to quickly retrieve the required information from the underlying data infrastructure is often complicated and takes a long time to program.
The Query Pilot
Our Solution
To solve the problem, we use the Pandata Query Pilot, a tool we developed to empower decision makers in making informed choices driven by data. With the Query Pilot, employees can access the existing data infrastructure via chat and query the information they need.
What was the turnover in Austria for 2023? How many customers became aware of us via LinkedIn?
Processes and queries that used to take half an hour can now be completed in seconds and no longer require any technical expertise.
_gif.gif)
Large Language Models (LLM)
Our Technology
The Query Pilot is based on a large language model (LLM) that converts text input into SQL queries. The queries are executed automatically and the result is displayed in an intuitive user interface.
Data protection is 100% guaranteed. The language model has no direct access to the data. It only receives the schemas and variable names of the tables in the data infrastructure and uses these to create a query. This query is then automatically forwarded to our customer's servers and executed locally. This ensures that no data ever leaves the ecosystem. The underlying language model can be selected and customized depending on the application.