Work · AI
Asking a geoscience database questions in plain English
A proof of concept for the WA Department of Mines that put a graph database and Power BI natural-language AI over public petroleum, mining and seismic data. Users could type questions like "show me all wells deeper than 500 m".
- Organisation
- WA Department of Mines × DataDivers
- My role
- Lead developer
- When
- May 2022
- Naturallanguage queries over geoscience data
- Graphmodel of wells, surveys and areas
- Public geoscience datawells, logs, seismic and geological surveys
- Azure SQL Graphentities and relationships
- Power BI modelvisual layer
- NLP / AI visualsplain-English questions
The problem
The department wanted to explore graph databases and make its public online platform easier to navigate and query. The platform holds petroleum wells, mining, seismic surveys and other geological surveys.
I knew this platform from the inside: I had used it for my own PhD research, so I understood exactly what slowed researchers down.
What I built
- A graph data layer on Azure SQL Graph that models wells, logs, surveys and areas and their relationships.
- A Power BI visual layer on top.
- Natural-language and AI visuals inside Power BI, so users could ask plain-English questions:
- “Show me the areas of wells that have wireline logging, including density logs.”
- “Show me the seismic data that crosses this area.”
- “List all wells deeper than 500 m.”
Outcome
A successful pilot that showed government stakeholders how conversational, graph-based analytics could make public geoscience data easier to use. It was my first engagement at DataDivers.