Building a GIS agent with Felt and Databricks

Geospatial workflows can be complex. Querying your data to visualize it on a map can take hours of work and require lots of GIS expertise.

What if anyone on your team could just ask, and an agent handled the rest: connecting your data, analyzing it, and turning it into a map? In this session, Felt and Databricks build that agent live, starting with your data in Databricks and finishing with a production map app in Felt.

We'll show:

• How Lakebase (Postgres + PostGIS on a serverless lakehouse) works as a geospatial engine
• How MCP tools wire up a live connection between Lakebase and Felt
• How Agent Bricks orchestrates the whole workflow from a single request

What started as an internal tool for Felt's Solution Engineers is now something any team working with spatial data can use.

‍

Speakers

Jaime Sanchez

Director of Partnerships | Felt

Jaime leads Felt's go-to-market and product relationships with leading Data & AI service providers such as AWS, Snowflake and Databricks. Jaime brings a strong foundation in computer science and over 10 years of GTM experience in geospatial tech which helps drive our strategy of bringing cloud native GIS to our customer's infrastructure.

Michael Johns

Geospatial Specialist Leader | Databricks

Michael leads customer adoption of scalable data solutions across the Americas, leveraging the power of the Databricks Lakehouse Platform for geospatial analytics. As a founding member of Databricks’ Geospatial SME group, Michael provides thought leadership, authors technical content, and speaks at industry events, helping shape the platform’s strategic direction.