Back to resources

Announcing SambaNova GPT Banking

Coins with a line graph over them.

Announcement Summary:

  • SambaNova is introducing a new deep learning model, GPT Banking, now generally available for customers
  • GPT Banking is a state-of-the-art large language model, specifically created for banking organizations
  • GPT Banking delivers high accuracy across a wide range of language tasks including language processing, understanding, classification, generation, and even translation with a single model
  • GPT Banking is available only on SambaNova’s Dataflow-as-a-ServiceTM platform, enabling organizations to deploy in weeks, not years, and greatly reducing effort necessary from data science teams

Earlier this month, we shared the exciting announcement that SambaNova had achieved new world records for training time, throughput and accuracy for GPT, an impressive achievement that reflects the hard work and dedication of SambaNova’s world class engineering team. During that announcement we also shared that, as the world record performance leader, ultimately, these performance metrics don’t matter. Customers don’t care about lab experiments, they care about improving their business and accelerating time to value. 

With this goal in mind, we are excited to announce SambaNova GPT Banking, a new version of GPT specifically trained for banking enterprises, and delivered through Dataflow-as-a-Service, enabling customers to transform their business with deep learning in weeks.

For those who are not familiar, GPT models are part of a new class of deep learning capabilities known as large language models. Although GPT and other large language models fall within the realm of AI known as Natural Language Processing (NLP), their capabilities go way beyond ‘chatbots’. What makes GPT so impressive is the high degree of accuracy it is able to provide across a wide range of language tasks such as language processing, understanding, classification, generation, and even translation. In fact, even with little or no specific training on solving a task, GPT can often outperform dedicated machine learning models that were specifically built to solve particular tasks, and in some cases even surpasses human level performance.

SambaNova takes this accuracy one step further. Going beyond a traditional GPT model, SambaNova GPT Banking has been pre-trained and refined specifically for the needs of enterprise banking organizations, enabling SambaNova to focus on optimizing performance and accuracy on the tasks that these organizations care about most.

While these performance metrics are impressive, what customers ultimately care about most is how AI and deep learning will help them improve their business by optimizing their customer experience, improving operational efficiency and managing risk and compliance. GPT Banking is a highly flexible deep learning model and the possibilities are nearly limitless for the areas of banking that it can help improve. Here are a few use cases providing competitive advantages to customers:

  • Customer experience optimization
  • Risk assessment in contracts and other large documents
  • Customer lifecycle management
  • Automated portfolio strategy optimization

When we discuss these possibilities with our banking customers, their next question is often “How fast can we get started?”, and the aspect of SambaNova GPT Banking that we are most excited about is the speed with which customers can deploy this state-of-the-art capability. Delivered through our Dataflow-as-a-ServiceTM platform, customers are able to deploy GPT Banking in weeks, not years, and greatly reduce the required time and effort from data science teams to train and deploy.

At SambaNova, we believe that GPT Banking isn’t just a new type of deep learning model, but a foundational shift that will transform nearly every aspect of banking operations across customer experience, back office operations, and risk and compliance.

Ready to learn more?

If you get the transformative potential of GPT Banking and deep learning and how it either plays as a competitive advantage or threat to your organization, apply for the SambaNova AI Impact Assessment program to learn how to get started .

As part of the AI Impact Assessment, SambaNova’s world-class team of machine learning and banking industry experts work directly with leading enterprises to evaluate and recommend specific areas where new deep learning capabilities like GPT Banking can provide the greatest short and long term impact.

If you are ready to start transforming your business with deep learning and GPT Banking, I hope you will consider applying to the AI Impact Assessment program today.

Image of Marshall Choy, the Vice President of Product at SambaNova Systems, responsible for product management and go-to-market
Marshall Choy

Marshall is the Senior Vice President of Product at SambaNova Systems, responsible for product management and go-to-market.

Back to resources