June 2025 Summaries

At the Databricks Data + AI Summit 2025 in San Francisco, significant advancements in geospatial analytics and AI were unveiled, notably the introduction of Agent Bricks and native Spatial SQL, heralding a new era for cloud-native geospatial analytics. Agent Bricks, a transformative framework, allows enterprises to build intelligent agents with ease by using plain language descriptions, enhancing performance through auto-evaluation and domain-specific reasoning. Meanwhile, the debut of native Spatial SQL on the Databricks platform enables large-scale spatial analysis with the introduction of new data types and the H3 Spatial Indexing system, optimizing geospatial data processing and rendering. Additionally, the summit highlighted the importance of interoperability with open data formats like Apache Iceberg, GeoParquet, and GeoArrow, enhancing the efficiency and integration of spatial analytics. These developments underscore a future where geospatial data is seamlessly integrated and AI is more accessible, positioning Databricks and CARTO at the forefront of this rapidly evolving landscape.

The blog discusses the integration of Google's Population Dynamics Foundation Model (PDFM) embeddings into CARTO Workflows, enhancing spatial analytics without requiring coding skills. It presents two use cases demonstrating how these embeddings improve predictions, specifically total liquor sales in Iowa and product-specific sales of Hawkeye Vodka. The integration allows users to leverage high-dimensional spatial representations that capture behavioral, demographic, and functional patterns, offering an alternative to traditional sociodemographic data. The results show that combining PDFM embeddings with demographic data yields the most accurate predictions, highlighting the embeddings' ability to capture latent spatial patterns not easily discernible through conventional data sources. This development aims to democratize access to advanced geospatial modeling, making spatial AI more accessible and impactful for various industries.

At the Snowflake Summit 2025, held in San Francisco, significant advancements were announced in the realm of geospatial data analytics, marking a pivotal shift towards cloud-native solutions. The event highlighted two major themes: interoperability and artificial intelligence (AI), both of which are redefining the use of spatial data. Interoperability was emphasized through the introduction of native geospatial support in Apache Iceberg and the adoption of GeoParquet as the standard for spatial data, enhancing performance and compatibility across platforms. On the AI front, Snowflake introduced Cortex, which includes AI SQL capabilities for querying unstructured data and agentic workflows for natural language interaction with spatial datasets. CARTO played a crucial role in these developments, integrating seamlessly with Snowflake to offer scalable and efficient spatial analytics. These advancements signify that geospatial data is becoming more accessible, scalable, and integral to broader data ecosystems.

CARTO's expert guide focuses on optimizing site planning strategies using data, AI, and real-time insights, addressing common challenges faced by organizations dealing with physical locations. It highlights six key strategies for enhancing site planning, including overcoming analytics skill barriers, embracing AI-driven decision-making, scaling decision-making processes, utilizing real-time data, optimizing network performance, and automating analytics workflows. The guide emphasizes the importance of integrating spatial analysis into decision-making to improve site lifecycle management and suggests that modern tools like CARTO's spatial analytics and AI agents can empower organizations to make faster and more confident location decisions. By providing flexible, intuitive self-service solutions and automating processes, organizations can reduce delays and increase the adoption of analytical tools, ultimately enhancing their location strategies across various industries.

Google's Population Dynamics Foundation Model (PDFM) is transforming the field of spatial analytics by learning from vast, multimodal datasets to represent how people interact with places over time. Unlike traditional models that are built for specific tasks using limited data, PDFM is trained on massive, diverse datasets to learn a deep, general-purpose understanding of human behavior in space. This shift enables the creation of rich embeddings that can be reused across many different applications, from disaster response to infrastructure planning. CARTO has explored the application of PDFM by predicting its own flagship spatial feature, the Human Activity Index, and found strong correlations between the model's outputs and expert-engineered indicators, demonstrating its real-world relevance. To further advance these models, CARTO is working on advancing them in areas such as moving beyond administrative boundaries, smart data selection, fine-tuning for local use cases, transparent and explainable AI, and coupling with agentic workflows.

Geo for the Agentic Era: Highlights from #SDSC25 London showcased a new era of GIS where AI agents and open data formats are transforming geospatial analysis, making spatial intelligence accessible to anyone. Leaders gathered to share their takes on challenges and opportunities in this reinvented field. Key highlights include the introduction of CARTO AI Agents, enabling natural language interaction with data, and the importance of open table formats like GeoParquet and Apache Iceberg for standardized storage and multi-engine access. The conference also emphasized the need for common frameworks to enable seamless collaboration across industries and highlighted the role of agentic AI in amplifying geospatial expertise rather than replacing it.