TopN AnalyticsNatural Language Generative BI SQL Agent
Client wanted a natural language interface for line managers to get quick data insights from database without having to depend on BI team to write and execute SQL queries
Where things stood
Non-Tech line managers and business leaders frequently needed quick insights from data which wasn’t always covered in standard dashboards.
To get this data, they relied on BI team members to write SQL queries and fetch the data from database tables.
This increased the turn-around-time (TAT) even for simple requests, leading to slower decision making within the organization.
What we built
We identified and integrated relevant data tables from the database into AI tool.
We created various aggregated tables for reducing complexity and improving response times.
We established table relationships, schema explanations, and sample queries for frequent use-cases. These were fed into AI’s memory.
How it works
The generative AI solution allowed users to ask questions in natural language such as “What’s yesterday’s revenue in Bangalore?” and the AI tool identified the correct table, formulated the correct SQL and responded with a data-backed answer. The solution showed the exact data that could be downloaded for further analysis along with a natural language response to the question asked.
The tool also provided graphical representation of the data in the form of bar charts, line charts or heatmaps.
What it's built on
What we'd do again
Creating pre-cleaned and aggregated tables significantly improved the AI model performance.
Want the full picture?
We're happy to walk through the details, numbers, and trade-offs directly.