What challenges did Quantum Capital Group face before AI implementation?

Manual geospatial planning for a single 20,000-acre development scenario took roughly three weeks. Comparing scenarios side by side took about six weeks, limiting how many configurations engineers could evaluate. The business needed faster planning across thousands of possible development configurations while respecting terrain and land-use constraints.

What was the impact of the AI solution?

Field development planning was reduced from about three weeks to around 20 minutes per scenario. Engineers can now run dozens of development scenarios per day instead of about one per month. The tool has analyzed 2.3 million surface pad candidates against 3.5 million subsurface well possibilities, covering more than 50 million permutations. Land utilization improved from about 50% to 85%, recovering 35 percentage points of leasehold that manual planning left undeveloped. Algorithmic improvements reduced solve times by 50% to 65% in an updated version of the tool.

Source

MicrosoftProductionEvidence: Medium65/100

Quantum Capital Group: Copilot Studio embedded agent for AI geospatial planning

Use case typePredictive decision supportUpdated Jun 13, 2026

Quantum Capital Group supports a portfolio company operating in the Piceance Basin of western Colorado, where energy exploration and field development planning are constrained by difficult terrain, land restrictions, and existing infrastructure. The company replaced a manual geospatial planning workflow that took about three weeks for a single development scenario and about six weeks to compare scenarios with an automated AI-driven tool built on Microsoft Azure. The tool uses an embedded Copilot Studio agent as a conversational interface for engineers and business stakeholders.

Organization: Quantum Capital Group
Industry: Energy & Utilities
Location: United States
Published: May 2026

Reported outcomes

50-65%

timeTime & speed

20 minutestime50-85%quantified impact

Strategic outcomes

Speed & agilityAccelerated field development planningScale & capacityEnabled many more scenarios per dayNew product / capabilityAdded conversational scenario analysisCustomer experience & trustImproved land utilization outcomes

Catalog median for time & speed deployments: −50% across 312 reported metrics. Compare benchmarks →

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Time: 20 minutes decrease

Microsoft Customer StoriesMay 12, 2026Customer storyInferred claimMedium evidence strength

Field development planning was reduced from about three weeks to around 20 minutes per scenario.

Normalized claim

Quantified impact: 50-85% increase

Microsoft Customer StoriesMay 12, 2026Customer storyInferred claimMedium evidence strength

Land utilization improved from about 50% to 85%, recovering 35 percentage points of leasehold that manual planning left undeveloped.

Normalized claim

Time: 50-65% decrease

Microsoft Customer StoriesMay 12, 2026Customer storyInferred claimMedium evidence strength

Algorithmic improvements reduced solve times by 50% to 65% in an updated version of the tool.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Quantum Capital Group
Provider: Microsoft
Maturity: Production
Linked source: Microsoft Customer Stories

The application uses Apache Sedona on Azure Databricks for geospatial analysis, Azure Blob Storage as the central data repository, and Dataverse as the operational data layer

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Geospatial optimization
2Conversational analytics
3Decision support

Manual geospatial planning for a single 20,000-acre development scenario took roughly three weeks.
Comparing scenarios side by side took about six weeks, limiting how many configurations engineers could evaluate.
The business needed faster planning across thousands of possible development configurations while respecting terrain and land-use constraints.

Quantum and its portfolio company built an automated field-planning tool on Microsoft Azure.
The application uses Apache Sedona on Azure Databricks for geospatial analysis, Azure Blob Storage as the central data repository, and Dataverse as the operational data layer.
The web application is hosted on Azure App Service and uses integer linear programming to select the best combination of pads and wells from a candidate pool.
A Copilot Studio agent is embedded in the web app to answer natural-language questions, explain optimization results, and generate plain-language scenario comparisons.

Technologies

Azure Databricks Azure App Service Copilot Studio Azure Blob Storage Dataverse Apache Sedona

Field development planning was reduced from about three weeks to around 20 minutes per scenario.
Engineers can now run dozens of development scenarios per day instead of about one per month.
The tool has analyzed 2.3 million surface pad candidates against 3.5 million subsurface well possibilities, covering more than 50 million permutations.
Land utilization improved from about 50% to 85%, recovering 35 percentage points of leasehold that manual planning left undeveloped.
Algorithmic improvements reduced solve times by 50% to 65% in an updated version of the tool.

Architecture

The solution uses Apache Sedona on Azure Databricks for geospatial optimization, Azure Blob Storage for planning data, Dataverse as the operational data layer, Azure App Service for the web front end, and an embedded Copilot Studio agent for conversational access to optimization results and scenario comparisons.

Sources & evidence1

Evidence: Medium65/100Evidence strength