Financial services organizations have large volumes of customer data, including account balances, payment transactions, and information such as customer FICO scores and credit history. Organizations are increasingly using cloud-based, GPU-accelerated artificial intelligence (AI) and machine learning (ML) predictive analytics recommender systems to make personalized suggestions to customers. However, building and maintaining recommender systems is a complex and time-consuming task. As technology partners, Microsoft and NVIDIA have custom tools to help develop and maintain recommendation systems.
What is a recommender system?
Recommender systems, also known as recommendation engines, are AI systems used to suggest a product, service or information to a user. Recommender systems are based on user characteristics, preferences, history and data, so the recommendation is always personalized for a particular customer or user.
Creating and maintaining recommender systems is complex
Historically, developing and maintaining recommender systems requires financial services personnel with special skills, such as data scientists or developers. Finding and maintaining the right recommendation algorithms can be a daunting task. Many financial organizations have legacy infrastructure, limited budgets for AI development, and staff that lack the data science skills needed to implement AI recommendation algorithms. This Forrester report study shows that “About two-thirds (64%) of technical decision makers are not fully confident in their ability to meet their organization’s AI goals based on current resources.”
A number of tasks are required to configure and test a recommender system ML model to meet the specific needs of an organization. These tasks include preparing data, building or selecting a recommendation algorithm model, tuning, training to optimize the model, and finally implementing the model. A first step is to collect, retrieve and organize data about customers and the financial products or services they use. Once the data is located, it must be brought together in a standardized format for use in AI or ML algorithms.
There are a number of existing recommendation algorithms available in repositories such as GitHub. As described in this Microsoft article“When asked to create a recommender system, data scientists often look to better-known algorithms to reduce the time and cost of choosing and testing more cutting-edge algorithms. Selecting the right algorithm Recommendering from scratch and implementing new models for recommender systems can be expensive as they require a lot of time for training and testing as well as large amounts of computing power.
Build an effective AI recommendation solution using GPU-accelerated cloud solutions
Training ML recommendation models requires huge computational resources. Legacy infrastructure with CPU-based processing cannot handle the required processing speeds. Moving to a GPU-based infrastructure allows for much faster processing and training for ML inference models and can help increase an organization’s return on investment (ROI).
According the Forrester investigation“What enterprises need are prebuilt and configurable AI cloud services. Cloud AI services allow developers to access a depth of AI capabilities through APIs to fuel application innovation without require data science experience Moving to a cloud-based AI solution that includes pre-built AI models, results in faster deployment time and gives organizations access to models AI tools that have been responsibly built and tested.
Using GPU-accelerated, cloud-based AI and ML solutions removes the barriers that financial services institutions face in developing AI and ML recommendation algorithms. NVIDIA”Survey on the state of AI in financial servicesfound that “enterprises derive significant financial benefits from enabling AI across the enterprise. More than 30% of respondents said AI increased annual revenue by more than 10%, while more than 25% said AI reduced annual costs by more than 10%.
Technology partners provide tools to help develop GPU-accelerated, cloud-based AI recommendation solutions
Microsoft and NVIDIA have a long history of collaborating and providing technology to help financial institutions build and implement AI recommendation systems. Using Microsoft Azure cloud, and the AI NVIDIA The platform provides the scalable and accelerated resources needed to run AI/ML algorithms, routines and libraries.
The partnership between Microsoft and NVIDIA makes powerful GPU acceleration available to financial institutions. The Azure Machine Learning Service integrates open-source NVIDIA QUICK software library that allows machine learning users to accelerate their pipelines with NVIDIA GPUs. NVIDIA TensorRT acceleration library has been added to ONNX Execution to accelerate deep learning inference. Azure supports NVIDIA T4 Tensor Core Graphics Processing Units (GPUs)and the NVIDIA DGX-H100 that are optimized for the cost-effective deployment of machine learning analytic or inference workloads.
NVIDIA Merlin framework designed for recommendation workflows
NVIDIA Merlin provides tools for building high-performance recommender systems at scale. Merlin includes libraries, methods and tools that streamline the construction of recommender systems. Merlin components and functionality are optimized to support retrieval, filtering, scoring, and ranking of hundreds of terabytes of data, all accessible through APIs. NVIDIA Merlin’s open source components make it easy to build and deploy a production-grade recommender system pipeline. Capital One has developed an industry-leading personalized recommendation architecture powered by NVIDIA’s ALBERT Transformer and Merlin Transformers4Rec algorithm that achieves superior performance by delivering relevant advertisements to repeat visitors to Capital One’s homepage. Sign up for NVIDIA GTC to learn more. https://www.nvidia.com/gtc/
Microsoft cloud-based solutions for financial recommendation systems
Move to the Microsoft Azure Cloud Solution provides financial institutions with a comprehensive set of computing, networking and storage resources integrated with workload services capable of handling the demands of recommendation algorithm processing. Microsoft Azure enables developers to build and train new AI models faster with automated machine learning, auto-scaling cloud computing, and built-in DevOps.
To help you create and implement recommendation algorithms, Microsoft provides a GitHub repository with examples of Python best practices to facilitate construction and evaluation of recommender systems using Azure Machine Learning services.
Financial services organizations are increasingly implementing AI recommendation systems to personalize product or service offerings to individual customers. Predictive analytics supported by cloud-based GPU-accelerated AI and ML recommender systems can improve customer experience and provide a new revenue stream for the organization.
But developing an ML recommender system is time-consuming and complex and requires personnel with specialized skills such as data science or programming. Microsoft and NVIDIA provide tools to streamline the process of developing, testing, training, and implementing recommender systems.