In addition to offering significant accessibility advantages for users with vision impairments, speech recognition solutions afford companies with increases to productivity while exceeding customer service expectations.
Speech can be captured faster than typing, with considerations given for interpreting native languages, tone of voice, and context. In addition, data is recorded, analyzed, and stored in real-time, providing ongoing machine learning (ML) and artificial intelligence (AI) potential.
See the following five case studies to learn how a variety of organizations across different industries are using speech recognition solutions:
1. Red Box
Originally founded in 1988, Red Box designed a digital recorder device that grew to dominate the industry. This strong foundation helped Red Box to become the world’s leading dedicated voice specialists, capable of capturing all voice communications across global enterprises.
With the need to provide high-quality transcripts for their customers, Red Box needed a platform that could be customized and scaled as needed. After evaluating a variety of speech recognition solutions, Red Box chose Deepgram for their accuracy, speed, and flexibility.
“We pride ourselves on how fast we can capture audio and deliver it to any best-of-breed analytics application, and Deepgram had a similar mentality when it came to the speed and high quality of its transcriptions,” says Pete Ellis, chief product officer, Red Box.
Industry: Software and telecommunications
Speech recognition solution: Deepgram’s AI Speech Platform
- Transcription accuracy improved to over 90%
- ASR transcription speed increased
- Customizable transcriptions are able to meet their customers’ needs
- Computing footprint was reduced by 7x
- Flexibility was available for on-premises transcriptions
As a leader in the development of integrated solutions for traceability and supply chain management, KFI relies on innovative solutions that increase the efficiency of processes. Using Vivoka’s Pick-by-Voice solution, KFI can vocalize a maximum number of tasks performed by their operators.
With a headset and nomad device configured for each user, Vivoka takes instructions and ensures their execution through the site.
Industry: Supply chain management
Speech recognition solution: Vivoka Pick-by-Voice
- Productivity losses were minimized
- Operator safety improved due to hands-free and eye-free operations
- Faster delivery of perishable products to points of sale was possible
- Precise management of selection and sorting of goods by destination and customer
- Monitoring of control points was simple and intuitive
- Instant data collection was possible
- Automated returns management, checking the condition of products, and quickly replacing them back into inventory
With 30 years of experience developing custom ERP and CRM software solutions, HansaWorld wanted to build on their reputation for building innovative products that help clients automate their business processes.
Building a talkbot using IBM technologies allowed HansaWorld to convert the audio streams provided by users into text, which is then analyzed, and an appropriate response or action is delivered back. Taking the solution to the next level is accomplished with IBM’s Watson Tone Analyzer, detecting the moods of users, and responding with consideration for frustrations and other emotions.
Industry: Software development
Speech recognition solutions: IBM Watson Speech to Text, IBM Watson Assistant, IBM Watson Text to Speech, and IBM Watson Tone Analyzer
- Talkbot service offers significant value to small and large clients
- More customer-friendly alternative to traditional hard-coded interactive voice systems
- Hundreds of hours are saved on administrative tasks
4. State Auto
Providing auto, home, and commercial insurance through independent agents and agencies in 33 U.S. states has made State Auto a major player in the industry. Protecting assets worth over $4.6 billion, State Auto needed an agile and flexible speech recognition solution that would make addressing customer needs faster and less costly.
Adding Amazon Transcribe to their existing Amazon technologies stack allows State Auto to derive insights from millions of calls made to customer service representatives (CSRs), converting speech to text quickly and accurately.
“Transcription is step one for anticipating the needs of customers who call us, and we build on top of it,” says Uthra Ramanujam, VP of strategic technology research, State Auto.
Speech recognition solution: Amazon Transcribe
- 15 ML models were built from the insights and data accrued
- Web application improved to better address customers queries
- Easier for customers to find requisite information online instead of calling support
- Reduced time and effort needed to assess the previous volume of over 5,000 claim calls per week
- Mitigated $800,000 in service operating costs
- Reviewed 83% more claims for potential fraud with detection three days earlier
5. The Ottawa Hospital
The Ottawa Hospital is a nonprofit, academic health sciences center located in Ottawa, Ontario. With 1,149 beds, The Ottawa Hospital serves 1.3 million people located across Eastern Ontario. With the mandate to improve patient safety and quality of care, driven by immediate access to accurate patient records, The Ottawa Hospital turned to Nuance Healthcare Solutions and their Dragon Medical Network Edition speech recognition solution.
Benefits like having notes from physicians available to nurses, clinicians, and surgeons instantly, instead of waiting two weeks, provides significant improvements to continuity of care. As the service standards are exceeded, physician liability is reduced while care is delivered consistently and accurately.
“We expect to save $11 million over five years because of Dragon Medical. The accessibility of patient records — possible because our clinicians are documenting care electronically — is truly exciting,” says Dr. Glen Geiger, chief medical information officer, The Ottawa Hospital.
Industry: Health care
Speech recognition solutions: Nuance Healthcare Solutions and Dragon Medical Network Edition
- Achieved $7 million in savings for transcription
- Documents created per month increased from 3,000 to 37,000
- Documentation is created two weeks faster and accessible through the circle of care
- Document accuracy, thoroughness, and consistency improved with the use of templates