Skip navigation EPAM
Dark Mode
Light Mode
CONTACT US

Optimizing HPC for the Cloud

How EPAM Helped SEFE Reduce Costs with Microsoft Azure Migration

At a glance

CLIENT

SEFE, London

STRATEGIC PARTNER

  • Microsoft Azure

SERVICES

  • Modernization
  • Cloud
  • Digital Risk Management

SEFE, a dynamic midstream energy company engaged in energy trading, portfolio management, sales and storage, embarked on a strategic initiative to modernize its high-performance computing (HPC) platform by migrating to Microsoft Azure. For this critical project, SEFE partnered with EPAM, a trusted collaborator since 2015, to navigate the complexities and enhance its technological capabilities.

THE NEED FOR SPEED

HPC in Commodities & Trading

SEFE relies on its HPC platform to produce intraday and end-of-day valuations for the derivatives desk and to provide risk teams with timely credit and market risk assessments. Shifting to Microsoft Azure aligns with SEFE’s strategy for the next generation trading landscape while delivering significant cost reductions. With a reference architecture provided by SEFE, EPAM used its deep engineering experience to deliver the client’s vision of a cloud native HPC solution called High Performance Compute Capability (HPCC). As the 2023 Microsoft US Azure Migrate Partner of the Year, EPAM demonstrated a deep understanding of leveraging Microsoft solutions to simplify cloud complexity and drive business value. SEFE knew that Azure and EPAM possessed the right technology and expertise to ensure success on its cloud journey.

Making the Shift to the Cloud

In 2020, SEFE recognized that its on-premises HPC platform’s software was reaching the end of its lifecycle. Looking for a cloud-based alternative, SEFE aimed to leverage the benefits of cloud computing, including increased capacity, scalability and cost-efficiency.

Drivers & Goals for Migration

SEFE established several key goals for its HPC migration:

Image
Reduce the total cost of ownership of on-premises hardware and applicable grid compute capability while providing transparent cost reporting and controls
Image
Enhance processing capacity, scalability and agility to support dynamic business requirements in a volatile market
Image
Enable rapid response to changes in business processing needs and support market analytics and reporting

EMBRACING THE CLOUD

A Modern, Optimized
HPC Platform

EPAM’s team of engineers, architects and cloud experts collaborated closely with SEFE to transition its legacy on-premises platform to native Microsoft Azure services. This innovative solution capitalized on Platform-as-a-Service (PaaS) features offered by the Azure platform, providing a technically robust and scalable alternative. Boasting an HPC cluster of 500+ on-demand servers, this solution offered unparalleled scale and agility, along with comprehensive monitoring for superior metrics and auditing compared to the legacy system.

​​With the new solution in place, SEFE not only reduced costs but also improved the speed and accuracy of decision making, setting the company apart in the energy trading industry.

Results

Image
Increased Scalability & Responsiveness
Image
Accelerated Time to Market
Image
Cost Savings Estimated at 66% Over the Next Five Years
Image
Better Financial Transparency
Image
Enhanced Resiliency

Looking to the Future

EPAM remains committed to supporting SEFE in its cloud transformation, with plans to migrate hundreds of data and computing servers to the cloud in the coming years.

HEAR FROM OUR CUSTOMERS

“EPAM has already proven its technical and delivery prowess over the last seven years with us and because of this successful engagement, SEFE has already realized multi-millions in hardware cost savings alone. EPAM has helped define a blueprint to a cloud-first, agile and modern trading IT landscape that supports a next generation trading business.”

Cuan Brown
EVP of Trading IT, SEFE Marketing & Trading

TECH STACK 

LEARN MORE

EPAM is the trusted Microsoft 
Solutions Partner for cloud mastery.


Explore more about our 
Microsoft expertise.