An overwhelming 91% of Financial Services Industry (FSI) companies evaluate or have already implemented artificial intelligence as a tool to drive innovation, improve operational efficiency, and improve customer experience.
Generative AI powered by NVIDIA NIM microservices and accelerated computing helps organizations improve portfolio optimization, fraud detection, customer service, and risk management.
Some of the companies leveraging these technologies to power financial services applications include Ntropy, Contextual AI, and NayaOne. Both are members of NVIDIA’s Inception program for cutting-edge startups.
Securiti, a Silicon Valley-based startup that provides a centralized, intelligent platform for securely using data and generated AI, is using NVIDIA NIM to build AI-powered solutions for financial services. Building a pilot.
At Money20/20, a major fintech conference taking place in Las Vegas this week, the companies will discuss how their technology makes disparate and complex FSI data actionable for banks, fintechs, payment providers and other organizations. demonstrate how they can be turned into valuable insights and opportunities for advanced innovation.
Ntropy brings order to unstructured financial data
Based in New York, Ntropy helps remove various states of entropy (disorder, randomness, and uncertainty) from financial services workflows.
“Every time money moves from point A to point B, text is left behind on bank statements, PDF receipts, and other forms of transaction history,” said Naré Vardanyan, co-founder and CEO of Ntropy. ” states. “Traditionally, that unstructured data has been very difficult to clean up and use for financial applications.”
The company’s transaction enrichment application programming interfaces (APIs) standardize financial data from disparate sources and geographies, enabling financial services applications to create human-like applications in just milliseconds at 10,000x lower cost than traditional methods. It serves as a common language that allows you to understand every transaction with precision.
It is built on Llama 3 NVIDIA NIM microservices and NVIDIA Triton Inference Server running on NVIDIA H100 Tensor Core GPUs. By using Llama 3 NIM microservices, Ntropy achieved up to 20x improvement in large-scale language model (LLM) utilization and throughput compared to running native models.
air force basea leading procure-to-pay software platform provider, uses LLM and Ntropy data enrichers to enhance its transaction authorization process.
At Money20/20, Ntropy explains how to use its API to clean up your customers’ merchant data. This improves the accuracy of risk detection models and enhances fraud detection. This reduces both the reduction of erroneous transactions and lost revenue.
Another demo shows how an automated loan agent leverages the Ntropy API to analyze information on a bank’s website and generate relevant investment reports to speed up loan diversification and the decision-making process for users. I will.
Contextual AI advances FSI’s search enhancement generation
Based in Mountain View, California, Contextual AI provides a production-grade AI platform powered by Search Augmented Generation (RAG), ideal for building enterprise AI applications in knowledge-intensive FSI use cases.
“RAG is the answer to bringing enterprise AI to production,” said Douwe Kiela, CEO and co-founder of Contextual AI. “By leveraging NVIDIA technology and large-scale language models, the Contextual AI RAG 2.0 platform brings accurate, auditable AI to FSI companies looking to optimize operations and deliver products powered by new generative AI. We can provide it.”
The Contextual AI platform integrates the entire RAG pipeline, including extraction, retrieval, re-ranking, and generation, into a single, optimized system that can be deployed in minutes and further tailored and specialized based on customer needs. This significantly improves the accuracy of context-sensitive tasks. .
HSBC plans to use contextual AI to provide research insights and process guidance support through the capture and synthesis of relevant market outlooks, financial news, and business documents. Other financial institutions are also using Contextual AI pre-built applications for financial analysis, generating policy compliance reports, resolving queries for financial advice, and more.
For example, a user can ask, “What is the forecast for central bank interest rates through Q4 2025?” The Contextual AI platform provides simple, fact-based explanations and accurate answers, including citations to specific sections of information sources.
Contextual AI uses NVIDIA Triton Inference Server and the open source NVIDIA TensorRT-LLM library to accelerate and optimize LLM inference performance.
NayaOne provides a digital sandbox for financial services innovation
London-based NayaOne offers an AI sandbox that allows customers to safely test and validate AI applications before commercial deployment. Its technology platform enables financial institutions to create synthetic data and gives them access to hundreds of fintech marketplaces.
Customers can use the digital sandbox to benchmark their applications for fairness, transparency, accuracy, and other compliance measures to better ensure the best performance and integration success.
said Karan Jain, CEO of NayaOne. “The demand for AI-driven solutions in financial services is accelerating, and our collaboration with NVIDIA will enable institutions to harness the power of generative AI in a controlled and secure environment.” We are creating an ecosystem that allows financial institutions to prototype faster and more effectively, leading to real business transformation and growth initiatives.”
NayaOne’s AI Sandbox, powered by NVIDIA NIM microservices, enables customers to explore, experiment with, and more easily deploy optimized AI models. With NVIDIA-accelerated computing, NayaOne can reduce infrastructure costs by up to 40% compared to running large CPU-based models while processing large datasets used in fraud detection models. Speed up processing by up to 10x.
Digital Sandbox also uses the NVIDIA RAPIDS set of open source data science and AI libraries to accelerate fraud detection and prevention capabilities in funds transfer applications. The company will demonstrate its digital sandbox at the NVIDIA AI Pavilion at Money20/20.
Securiti improves financial planning with AI Copilot
Powering a wide range of generative AI applications, including secure enterprise AI co-pilot and LLM training and tuning, Securiti’s highly flexible Data+AI platform enables users to build secure end-to-end enterprise AI systems .
The company is currently building a financial planning assistant powered by NVIDIA NIM. Co-pilot chatbots access a variety of financial data and provide context-aware responses to users’ financial questions, while adhering to privacy and eligibility policies.
“Banks are struggling to deliver personalized financial advice at scale while maintaining data security, privacy, and regulatory compliance,” said Jack Berkowitz, Security’s chief data officer. Ta. “Securiti is helping build a secure AI co-pilot that delivers personalized financial advice tailored to individual goals, with robust data protection and role-based access for secure and scalable support. Masu.”
Chatbots pull data from a variety of sources, including revenue records, customer profiles and account balances, and investment research documents. Securiti’s solutions securely ingest data, prepare it for use with high-performance NVIDIA-powered LLM, and maintain control over access rights and more. Finally, provide customized responses to users through a simple consumer interface.
Securiti used Llama 3 70B-Instruct NIM microservices to optimize LLM performance while ensuring secure use of data. The company will demonstrate its generative AI solution at Money20/20.
NIM microservices and Triton Inference Server are available through the NVIDIA AI Enterprise software platform.
Join NVIDIA at Money20/20, running through Wednesday, October 30th, to learn more about AI for financial services.
Explore new NVIDIA AI workflows for fraud detection.
