Each colour maps a published value to the exact source passage used to support it. Only bounded excerpts are public; administrators can inspect the complete captured source.
Title
Discover Financial Services Builds a Generative AI Solution on AWS for Faster Decision-Making and Time to Market
Discover Financial Services Builds a Generative AI Solution on AWS for Faster Decision-Making and Time to Market Skip to main content Filter: All English Contact us AWS Marketplace Support My…
Company
Discover Financial Services
…ed speed up time to market, mitigate risk, and improve the customer experience. About Discover Financial Services Discover Financial Services is a leading digital banking and payment services c…
Description
Discover Financial Services, a digital banking and payment services company, built an analytics workbench and unified data science workbench on Amazon EC2 (P3 and P4 GPU instances) so its data scientists could run machine learning and generative AI workloads, train large language models with sample sizes requiring up to 6 TB of memory, and deliver high-performance computing in the cloud. Model artifacts are stored in Amazon S3 and shared across engineering teams via Amazon EFS. Discover used the platform for use cases including sentiment analysis of customer service calls and a 'do not contact' model that classifies customers in near real time for customer care agents. Using feature embedding, the team reduced time to market from hours to minutes; parallel model training cut processing of 30 million records from days to hours, and sentiment analysis on a 57,000-record dataset dropped from hours to minutes. Discover reports 35% reduction in engineering and platform costs.
…orkbench that its data scientists can use to run or process AI/ML applications, train models with large sample sizes—requiring up to 6 TB of memory—and deliver high-performance computing (HPC) in the cloud using core GPUs. “We have provided our scientists with a cloud-scale data warehouse with on-dem…
Problem
In its various lines of business, including decision and credit and portfolio risk management, Discover faced the challenge of making sure that it could launch its services fast enough. We wanted to make decisions quicker and get to insights faster so that we could respond faster to our customers.
…lines of business, including decision and credit and portfolio risk management, Discover faced the challenge of making sure that it could launch its services fast enough. “We wanted to make decisions quicker and get to insights faster so that we cou…
Implementation approach
The team ran benchmark tests using different models and codes to assess processing speed: using 16 CPUs, processing 20,000 rows took 6.5-7 hours, while a single-GPU setup took 23 minutes, and using multiple GPUs brought the processing time down to 4 minutes but also increased costs, helping the team choose different GPU setups based on use case requirements and cost.
…ribed from recorded interactions between customer service agents and customers. Using 16 CPUs, the processing took 6.5–7 hours, while a single-GPU setup took 23 minutes. Using multiple GPUs brought the processing time down to 4 minutes but also incr…
Industry
Financial Services
…s and accelerate the development of innovative renewable energy business models Financial Services Develop innovative and secure solutions across banking, capital markets, insurance, and payments Games Enable game development for every genre and platform, from AAA titles to…
Technology
Amazon EC2, Amazon S3, Amazon EFS
…rience. Rahul Gupta AI/ML Expert Platform Engineer, Discover Financial Services AWS Services Used Amazon EC2 P4 Instances Delivers high performance for machine learning (ML) training and high performan…
Business functions
Customer Service & Support
…y such customers. The solution classified these customers in near real time and made the relevant data available to customer care agents. The agents could thus identify customers who should not be contacted, which he…
AI capabilities
Generative AI, Predictive Analytics
…o create a data science solution that its scientists and analysts could use for machine learning (ML) and generative artificial intelligence (AI) workloads. The solution helped speed up time to market, mitigate risk, and improve the cu…
Deployment options
cloud
…ations, train models with large sample sizes—requiring up to 6 TB of memory—and deliver high-performance computing (HPC) in the cloud using core GPUs. “We have provided our scientists with a cloud-scale data warehouse with on-dem…
Country
United States
…ta Science Solution with Generative AI Capabilities for Reducing Time to Market Chicago-based Discover, a digital banking and payment services company, aims to help people spend smar…
