How KPMG Mines Insights for 200 Billion Data Rows
Learn how KPMG turned data overload into a strategic asset, boosting efficiency and client value.
With clients across 145 countries and a workforce of 236,000 people, KPMG found itself drowning in a sea of data, struggling to extract meaningful insights.
The company decided it was time to transform its data landscape and chose Microsoft Fabric to drive the change.
In today's edition, we’ll examine KPMG's specific challenges, how it overcame them, and the lessons we can learn.
The data dilemma: when more isn't always better
Rajiv Phougat, Principal and CTO for Data and Analytics at KPMG US, described the struggle as follows: "The challenge was more than stitching multiple components together. It was imperative that we quickly reduce the complexity of the footprint and deliver business value to our employees and clients."
Here's what KPMG was grappling with:
Volume overload and complexity: Data ranging from a few million to billions of rows daily with multiple sources and types
Lack of global access: The need for worldwide accessibility to data and insights
Insight drought: Difficulty in quickly moving from raw data to actionable insights
Efficiency emergency: High demands from IT personnel and slow data load times
The impact is a hampered ability to provide advanced services to clients. In an industry where timely insights can make or break multimillion-dollar decisions, KPMG needed a solution—fast.
The solution: Microsoft Fabric
KPMG chose Microsoft Fabric for its complex data challenge. But why?
(Source)
All-in-one analytics platform: From data movement to data science, analytics, and business intelligence, Fabric provides a comprehensive suite of services.
Cloud-powered agility: It offered KPMG the flexibility and scalability they needed to handle their global data demands.
Seamless integration: Microsoft Fabric seamlessly integrated various elements to create a complete analytics solution.
Let's look at the key components of Fabric that KPMG used:
OneLake
OneLake became KPMG's central data repository, allowing them to connect various sources regardless of location or storage mechanism.
This solved its global accessibility issues and reduced data silos. Phougat explains, "We can see them in one place in a single pane of glass, then provision with the proper permissions and security to make the data available to our internal consumers."
Power BI
KPMG used Power BI for data visualization and creating analytics, dashboards, and reports. With Copilot in Power BI, KPMG's developers can:
Generate narratives automatically
Create simple Q&A interfaces with prebuilt prompts
Expedite visualization development
Azure Databricks
Azure Databricks became KPMG's solution for complex data transformations and specific use case modeling. "Azure Databricks is our standard way of building any transformation or landing pipeline," Phougat explained. This allowed KPMG to process its billions of rows of data more efficiently.
Furthermore, KPMG developed its solutions on top of the Microsoft stack:
KymChat: An AI-powered chatbot used by thousands of KPMG professionals in Australia, built using Microsoft OpenAI Service.
KPMG powered enterprise data & analytics: A solution helping organizations accelerate their data and AI journey, powered by Microsoft Fabric.
Multi-cloud FinOps solution: Built using Microsoft Fabric to optimize cloud expenditure for KPMG firms and clients.Focused migration approach
The Impact
Here's what KPMG achieved using Microsoft Fabric:
Rapid analysis at scale: Teams can now quickly analyze vast amounts of data. For instance, audit teams reduced the time to process 200 billion rows of client data from weeks to days.
Improved client services: KPMG can offer deeper insights into operational efficiencies, market opportunities, and strategic planning.
Improved collaboration: Fabric's unified approach to data analytics has helped KPMG employees devise creative solutions to business challenges.
AI-powered efficiency: With Copilot and Azure OpenAI Service, KPMG can now generate agent-based interactive modeling and presentations in hours instead of days or weeks.
Streamlined operations: KPMG has optimized its internal processes by reducing data load times and easing IT personnel demands.
Lessons learned
KPMG's journey with Microsoft Fabric offers valuable lessons for any organization looking to change its data landscape:
Prioritize usability: Ensure your data tools are accessible to technical and business teams. Choose solutions that seamlessly connect different data sources and tools.
Use AI: Integrate AI and machine learning to accelerate analysis and deeper insights.
Balance tech and human expertise: While AI can do much, human judgment remains crucial. Aim for the right combination of automation and human insight.
Think beyond analysis: A truly transformative data solution should provide insights, improve collaboration, and spark creation.
Learn more here.
Here are some of the insightful editions you may have missed: