The audio transcription AI workflow uses the NVIDIA Riva English out-of-the-box automatic speech recognition (ASR) AI service, customizes it for the best possible real-time accuracy for the financial service industry use case, and deploys it integrated into a reference enterprise-ready solution.
Example training and inferencing pipelines are built using Riva to demonstrate how to customize and build a transcription solution for a specific use case, in this case financial services. Additional supporting components are also integrated as shown in the below diagram.
Key benefits of the Riva audio transcription AI workflow:
The Best Possible Accuracy: Achieve the best possible real-time English audio transcription accuracy by fine-tuning Riva ASR models for the finance service industry use case.
Seamless Scaling: Instantly scale to hundreds and thousands of English audio transcripts with microservices deployable on any cloud Kubernetes distribution.
Fast & Flexible Deployment: Quickly get started with a deployable English audio transcription in the cloud, on-premises, or on edge.
To get started, review the documentation linked below for more information on what is involved and included in the workflow, and how to deploy and run the workflow.
Join the NVIDIA Developer Program to access a 90-day free trial of NVIDIA Riva and access the audio transcription workflow through NGC.
Have an upcoming speech AI project? Try the transcription AI workflow on NVIDIA LaunchPad.
Contact NVIDIA to find out how you can purchase NVIDIA Riva for your production deployment.
NVIDIA Speech AI Audio Transcription Workflow Deployment Guide
NVIDIA Speech AI Audio Transcription Workflow Solution Brief
Use the NVIDIA Riva SDK to build your own speech- and translation-AI-based solutions.
Learn more about how to use NVIDIA Riva through our Deep Learning Institute platform. You can work with Riva interactively on hardware provided by NVIDIA hosted in the cloud.
You must have a developer account and be signed in to access the following Riva courses: