Leaders in Fintech Use Generative AI to Provide Faster, Safer, and More Accurate Financial Services.
Ntropy, Contextual AI, NayaOne, and Securiti improve financial planning, fraud detection, and other AI applications with NVIDIA NIM microservices and quicker processing. A staggering 91% of businesses in the financial services sector (FSI) are either evaluating artificial intelligence or currently using it as a tool to improve client experiences, increase operational efficiency, and spur innovation.
Generative AI powered by NVIDIA NIM microservices and quicker processing may improve risk management, fraud detection, portfolio optimization, and customer service.
Companies like Ntropy, Contextual AI, and NayaOne all part of the NVIDIA Inception program for innovative startups are using these technologies to improve financial services applications.
Additionally, NVIDIA NIM is being used by Silicon Valley-based firm Securiti to develop an AI-powered copilot for financial services. Securiti is a centralized, intelligent platform for data and generative AI safety.
The businesses will show how their technology can transform heterogeneous, sometimes complicated FSI data into actionable insights and enhanced innovation possibilities for banks, fintechs, payment providers, and other organizations at Money20/20, a premier fintech conference taking place this week in Las Vegas.
Ntropy Brings Order to Unstructured Financial Data
New York-based Ntropy Organizes Unstructured Financial Data Ntropy assists in clearing financial services processes of different entropy disorder, unpredictability, or uncertainty states.
By standardizing financial data from various sources and geographical locations, the company’s transaction enrichment application programming interface (API) serves as a common language that enables financial services applications to comprehend any transaction with human-like accuracy in milliseconds, at a 10,000x lower cost than conventional techniques.
The NVIDIA Triton Inference Server and Llama 3 NVIDIA NIM microservice use NVIDIA H100 Tensor Core GPUs. The Llama 3 NIM microservice increased Ntropy’s large language models (LLMs) usage and throughput by 20x compared to native models.
Using LLMs and the Ntropy data enricher, Airbase, a top supplier of procure-to-pay software platforms, improves transaction authorization procedures.
Ntropy will talk at Money20/20 about how their API may be used to clean up merchant data belonging to consumers, which increases fraud detection by enhancing risk-detection algorithms’ accuracy. Consequently, this lowers revenue loss and erroneous transaction declines.
In order to expedite loan distribution and user decision-making, an additional demonstration will demonstrate how an automated loan agent uses the Ntropy API to examine data on a bank’s website submit an appropriate investment report.
What Is A Contextual AI?
Contextual AI perceives and reacts to its surroundings. This implies that when it answers, it takes into account the user’s location, prior actions, and other crucial information. These systems are designed to provide customized and relevant responses.
Contextual AI Advances Retrieval-Augmented Generation for FSI
A California-based company with headquarters in Mountain View, provides a production-grade AI platform that is perfect for developing corporate AI applications in knowledge-intensive FSI use cases. Retriever-augmented generation powers this platform.
In order to provide significantly higher accuracy in context-dependent tasks, the Contextual AI platform combines the entire RAG pipeline extraction, retrieval, reranking, and generation into a single, optimized system that can be set up in a matter of minutes and further customized and tuned in response to user requirements.
HSBC intends to employ contextual AI to retrieve and synthesize pertinent market outlooks, financial news, and operational papers in order to enhance research findings and process recommendations. Contextual AI’s pre-built applications, which include financial analysis, policy-compliance report production, financial advising inquiry resolution, and more, are also being used by other financial institutions.
A user may inquire, “What’s our forecast for central bank rates by Q4 2025?” for instance. With references to certain parts of the source, the Contextual AI platform would provide a succinct explanation and a precise response based on real documents.
Contextual AI works with the open-source NVIDIA TensorRT-LLM library and NVIDIA Triton Inference Server to improve LLM inference performance.
NayaOne Provides Digital Sandbox for Financial Services Innovation
London-based NayaOne Offers a Digital Sandbox for Financial Services Innovation. Customers may safely test and certify AI applications using NayaOne‘s AI sandbox before they are commercially deployed. Financial institutions may develop synthetic data and access hundreds of fintechs on its platform.
Customers may utilize the digital sandbox to better assure top performance and effective integration by benchmarking apps for accuracy, fairness, transparency, and other compliance standards.
The need for AI-powered financial services solutions is growing, and our partnership with NVIDIA enables organizations to use generative AI’s potential in a safe, regulated setting. “Its’re building an ecosystem that will enable financial institutions to prototype more quickly and efficiently, resulting in genuine business transformation and expansion projects.”
Customers may explore and experiment with optimal AI models using NayaOne‘s AI Sandbox, which makes use of NVIDIA NIM microservices, and more quickly deploy them. When using NVIDIA accelerated computing, NayaOne can analyze massive datasets for its fraud detection models up to 10 times quicker and with up to 40% less infrastructure cost than when using extensive CPU-based algorithms.
Using the open-source NVIDIA RAPIDS data science and AI libraries, the digital sandbox speeds up money movement fraud detection and prevention. At the NVIDIA AI Pavilion at Money20/20, the company will display its digital sandbox.
Securiti’s AI Copilot Enhances Financial Planning
Securiti’s very adaptable Data+AI platform enables customers to create secure, end-to-end corporate AI systems, supporting a wide variety of generative AI applications such as safe enterprise AI copilots and LLM training and tuning.
The business is currently developing a financial planning assistant that is driven by NVIDIA NIM. In order to deliver context-aware answers to customers’ financial inquiries, the copilot chatbot accesses a variety of financial data while abiding by privacy and entitlement regulations.
The chatbot pulls information from a number of sources, including investing research materials, customer profiles and account balances, and earnings transcripts. Securiti’s technology preserves controls like access entitlements while securely ingesting and preparing information for usage with high-performance, NVIDIA-powered LLMs. Lastly, it offers consumers personalized replies via an easy-to-use user interface.
Securiti ensured the secure usage of data while optimizing the LLM’s performance using the Llama 3 70B-Instruct NIM microservice. The company will demonstrate generative AI at Money20/20. The NVIDIA AI Enterprise software platform offers Triton Inference Server and NIM microservices.
