How EY and Xoople bring geospatial intelligence into enterprise decision-making

The Earth is constantly sending signals. Vegetation stress weeks before a harvest shortfall. Urban growth patterns that reshape consumer demand and catastrophic flooding events that close distribution corridors before a logistics team has seen a weather alert. The data exists and the science is mature, but for most enterprises, these physical-world signals have remained out of reach because they are too complex to operationalize and too disconnected from the systems where supply chain decisions actually get made. 

EY and Xoople, building together on Microsoft, set out to close that gap. Combining EY’s deep supply chain consulting experience and enterprise client relationships with Xoople’s scientific-grade Earth data, grounded in the infrastructure of Microsoft Planetary Computer Pro, Fabric, and Foundry, the three companies have built a solution that brings geospatial intelligence into enterprise demand forecasting and scenario planning as a first-class data source that is accessible so that the right data is available at the right time.

The result is not proof of concept. It is an enterprise-ready capability, already in use with global clients, with a clear pathway to scale.

The challenge

Physical world data has long been one of the most underused assets in enterprise decision making. Not because of a lack of data (satellite coverage is now extraordinary), but because of two compounding barriers that have kept geospatial intelligence siloed from the businesses that need it.

The first barrier is operationalization. Ingesting and processing raw satellite imagery requires costly, bespoke infrastructure. Organizations that wanted to use geospatial data faced the prospect of standing up specialized pipelines, maintaining complex tooling, and building or hiring GIS expertise that sat entirely outside the enterprise data stack. Up to now, economics could rarely justify the effort.

The second barrier is the workflow gap. Even where organizations cleared the infrastructure hurdle, the output (geospatial insights interpreted by specialists) rarely made it into the core planning systems where it could change decisions. Demand models, inventory optimization engines, and scenario planning tools ran without visibility into the physical conditions that often determine outcomes, so the available intelligence existed in silos.

EY saw this gap acutely through its supply chain practice. Across manufacturing, retail, and logistics clients, demand models were running blind to environmental and physical signals that were directly material to forecast accuracy. EY recognized that closing this gap required more than better data; it required a fundamentally different delivery architecture.

The collaboration

The solution required three things working in combination: an enterprise-grade geospatial data substrate, a platform that could deliver that data directly into existing enterprise workflows, and the supply chain experience to connect the data to real business outcomes. EY, Xoople, and Microsoft each provide one of these three missing pieces.

What Xoople brings 

Xoople is the Earth data infrastructure layer built for AI, enabling it to understand and predict daily physical changes on the world’s surface. It serves businesses, organizations, and governments that need to know what is physically happening in the world in order to move faster, reduce risk, and make decisions grounded in reality, across supply chain and logistics, infrastructure monitoring, agriculture, insurance, urban planning, and security.

Xoople has developed technology that converts trillions of pixels and precise measurements into consumable, open formats. Not as satellite imagery that requires heavy pre-processing. Not as a separate dashboard to check. As a structured dataset that feeds directly into business workflows, AI models, and the agentic systems that act on that intelligence.

Xoople’s upcoming proprietary constellation, developed in collaboration with L3Harris, generates 100x more Earth data than the current legacy standard through a fully integrated space-to-cloud network, purpose-built so that when AI models and agents are making decisions at scale, accuracy is no longer a nice-to-have. Xoople’s founders spent decades designing some of the most significant Earth observation missions in history. They saw the same problem repeat: extraordinary data collected from space, and almost none of it reaching the organizations that needed it most. Xoople was built to close that gap, with a vision of Earth intelligence as an invisible infrastructure layer of the global economy, as ubiquitous and foundational as GPS.

What EY brings 

EY contributes four elements that transform geospatial data from a promising signal into a deployable enterprise capability:

What Microsoft brings

Microsoft has built an enterprise-grade geospatial data substrate. The combination of Planetary Computer Pro, Anyscale on Azure, Azure Kubernetes Service, Fabric, and Foundry removes the barriers that have historically kept geospatial data out of enterprise workflows:

• Planetary Computer Pro removes the infrastructure barriers that often limit the use of geospatial data in enterprise workflows. It provides a Microsoft-native home for Xoople’s AI-ready satellite data alongside curated geospatial and environmental datasets, as well as outputs from geospatial AI models hosted in Microsoft Foundry—all without requiring customers to build, maintain, or scale custom data pipelines. By connecting geospatial data with the broader enterprise data ecosystem, Planetary Computer Pro enables organizations to seamlessly use location intelligence within platforms such as Microsoft Fabric and Microsoft Foundry, accelerating analytics, AI development, and business decisionmaking.

• Anyscale on Azure allows Xoople to run distributed AI workloads by bringing the open-source compute framework, Ray, and the Anyscale Runtime natively to Azure, all running on Azure Kubernetes Service. 

• Microsoft Fabric removes the integration barrier, making geospatial data directly consumable in the Lakehouse alongside the enterprise tables clients already operate on.

• Microsoft Foundry and Aurora remove the workflow gap, providing the agentic and atmospheric modeling layer that converts measurement into action, and the chat interface through which enterprise users initiate scenario planning without technical mediation.

The Solution

The joint solution is an end-to-end geospatial intelligence platform for enterprise demand forecasting and scenario planning, built on a tightly integrated Microsoft stack.

At the foundation, Xoople’s Planetary Computer Pro instance provides the Earth data layer. Sentinel-2-derived geospatial variables, including vegetation indices, surface water coverage, and urban change signals, are curated alongside premium tabular and raster data and supply chain collections. Aurora, Microsoft’s foundation model of the atmosphere hosted in Foundry, adds a weather modeling dimension: atmospheric forecasts are generated using data sourced from Planetary Computer and outputs are directly ingested back to Enterprise's Planetary Computer Pro instance, providing an additional forward-looking signal for demand modeling.

Anyscale on Azure provides the distributed compute and runtime layer that operationalizes Xoople’s geospatial AI at scale, enabling Ray-based pipelines to process and model planetary-scale satellite data across Azure Kubernetes Service with high-performance, fault-tolerant execution that seamlessly integrates with Planetary Computer Pro and Foundry for end-to-end data-to-decision workflows.

“What Xoople is doing is incredibly important as they are creating Earth's System of Record, the data layer AI has been missing, whilst also defining a new category that will transform decisionmaking for the Enterprise,” says Xoople VP of Engineering, Milos Colic. "With Anyscale on Azure, Xoople can reliably run massively distributed AI workloads over planetary-scale satellite imagery, transforming complex spectral data into decision-ready intelligence. Anyscale lets our teams focus on models and outcomes rather than infrastructure, dramatically accelerating experimentation to deployment. For our product teams and theirs, this means a faster stream of information, more agility and improved risk management.”

EY Supply Chain Data is brought into the solution alongside Xoople’s geospatial variables, combined in a shared Enterprise Planetary Computer Pro instance and a Fabric Lakehouse. This unified data environment is where geospatial intelligence meets enterprise context.

A Demand Forecasting Notebook in Microsoft Fabric sits at the center of the solution. It combines satellite-derived geospatial variables with EY’s supply chain data to generate demand forecasts across a range of products and locations. These forecasts are surfaced to business users through a custom reporting dashboard hosted on Azure App Service, enabling location- and product-level demand views without requiring specialized geospatial expertise to operate.

For scenario planning, users access a chat interface that triggers Foundry-deployed agentic workflows. Foundry agents query enterprise's own Planetary Computer Pro instance and Fabric Lakehouse via custom tools deployed through Azure Functions and return structured scenario analyses. The result is a system where enterprise users can initiate complex geospatial scenario planning through natural language, designed to significantly reduce infrastructure complexity for enterprise users.

The impact

Initial results from this deployment demonstrate the potential of both the technical approach and the business case.

In a demand forecasting engagement at a global food and beverage company, an enterprise data scientist was given access to Xoople’s geospatial data product to integrate into existing forecasting workflows. Xoople’s geospatial variables provided a meaningful signal for approximately two-thirds of the SKUs evaluated. For a demand model seeking every marginal improvement in forecast accuracy, this represents a material new input.

Beyond the individual result, the significance is structural. Together, Xoople’s AI-ready geospatial data and Planetary Computer Pro’s operational geospatial platform dramatically reduce the data engineering, compute overhead, and bespoke pipelines traditionally required to work with Earth observation data. By delivering geospatial signals in enterprise-ready tabular formats aligned with existing business and geographic hierarchies, the solution enables data scientists and analytics teams to integrate Earth data directly into their existing AI and analytics workflows—without specialized GIS infrastructure or major changes to the systems already in place.

Supply chain is the initial enterprise focus, representing a highly data-driven function with growing AI adoption. But the same delivery pattern, Earth data in enterprise-ready tabular formats integrated directly into existing analytics platforms, applies across agriculture, insurance, commodity trading, utilities, property and catastrophe, forestry, infrastructure, and water. With the EY-Xoople-Microsoft solution proven in supply chain, the pathway to adjacent industries is clear.

Shared perspectives

EY perspective
EY brings to this solution the supply chain consulting expertise, proprietary EY supply chain data, and the enterprise client relationships that translate geospatial capability into business outcomes. As supply chain solution lead, EY has designed the end-to-end integration that connects Xoople’s Earth data to the forecasting and scenario planning workflows enterprise clients rely on every day. 

“Supply chain planning has traditionally operated with limited visibility into real-world conditions," says Dany Bou, AI & Data Supply Chain Leader at EY. "By combining EY’s supply chain expertise, Xoople’s geospatial intelligence, and Microsoft technologies including Planetary Computer Pro, Fabric, and Foundry, we are helping clients bring physical-world signals directly into enterprise planning workflows and make faster, more informed operational decisions.”

Xoople perspective
Xoople’s goal was always to make Earth intelligence as accessible as any other dataset, something a supply chain director or a CFO could act on, not just a GIS expert. Microsoft answered the computational needs: Xoople runs on Azure, deeply integrated with Microsoft Planetary Computer Pro, so its intelligence is distributed through the cloud environments where enterprise buyers already operate. The integration with Microsoft Foundry takes this further. Foundry’s agentic capabilities mean Xoople’s dataset is not just available for human analysts; it becomes the ground-truth layer that AI agents can reason over and act on. As enterprises move toward agentic workflows that take action without human intervention, those agents need reliable data about what is physically happening in the world. Xoople provides that ground truth. 

“The next frontier for enterprise AI isn’t a better model, it’s better ground truth," explains Fabrizio Pirondini, CEO of Xoople. "Through Microsoft Planetary Computer Pro, Fabric, and Foundry, Xoople’s Earth data flows directly into the supply chain, forecasting, and risk workflows where decisions are made. What we’re building with Microsoft and EY is proof that physical-world intelligence is no longer a specialist capability. It’s enterprise data infrastructure.”

Looking ahead

The foundation is in place. The solution has been demonstrated in initial enterprise engagements. And the scale opportunity is significant.

With EY, Xoople is extending the same pattern across a broader set of enterprise clients, with active build-out already underway. The delivery architecture, Xoople's enterprise-ready Earth data from Planetary Computer Pro integrated directly into Fabric alongside existing enterprise data with scenario planning powered by Foundry agents, is replicable across the supply chain client base without rebuilding from scratch for each deployment.

Looking further ahead, the launch of Xoople’s constellation will transform the resolution and density of geospatial signals available to enterprise clients. A projected improvement of up to 100x in data density, on the same delivery pattern enterprises are already consuming today, means demand models will have access to finer-grained physical-world signals (more SKUs, more locations, more conditions) without any change to the integration architecture.

The broader ambition is clear: to expand geospatial intelligence from the 12 million traditional users¹ who have historically operated it as a specialist discipline, to the more than 223 million global AI users¹, data scientists, and BI developers who already work in the platforms where this data can now land. EY’s supply chain practice, Xoople’s Earth data, and Microsoft’s platform are the three-way alignment that makes that expansion possible.

1. 12M geospatial users as calculated from ESRI users and QGIS users. 223M data users is a composite statistic from the following sources: ChatGPT users (Nov 24), Power Platform users, PowerBI developers, and global data science jobs per Fortune Business. 

2. 450M people interacting with supply chain and 75% of supply chain professionals using AI-powered analytics.

About EY
EY exists to build a better working world, helping to create long-term value for clients, people and society and build trust in the capital markets. Enabled by data and technology, diverse EY teams in over 150 countries provide trust through assurance and help clients grow, transform, and operate. In this relationship, EY contributes its AI & Data practice and supply chain consulting experience, acting as systems integrator and supply chain solution lead.

About Xoople
Xoople is the Earth data infrastructure layer built for AI, enabling enterprises, governments, and organizations to understand and predict daily physical changes on the world’s surface. Founded in 2019 and headquartered in Tres Cantos, Madrid, Spain, Xoople converts satellite imagery into consumable, open tabular formats that feed directly into business workflows, AI models, and agentic systems. Through its upcoming constellation developed with L3Harris and deep integration with Microsoft Planetary Computer Pro, Xoople is building Earth intelligence purpose-built for the AI era, with 100x more data than the current legacy standard. Xoople serves clients across supply chain and logistics, infrastructure monitoring, agriculture, insurance, urban planning, and security.