Caption Companion Featured on Assistive Technology Update (ATU735)
Premier Visual Voice was recently featured on Assistive Technology Update, a nationally recognized podcast produced by Easterseals Crossroads and hosted by Josh Anderson. In Episode 735 (ATU735), Executive Director Margaret Downs and Director of Technology Mathew Call joined the show to discuss the evolution of Caption Companion, recent certifications, and why offline, audio-first captioning is redefining accessibility.
This episode offered a deep, practical look at how Caption Companion was built, how it performs in real-world environments, and why it has become a trusted assistive technology for individuals, schools, libraries, healthcare settings, and public venues.
Below is a comprehensive summary of the discussion, including the key questions asked during the episode and the insights shared by the Caption Companion team.
Assistive Technology Update (ATU) is a long-running weekly podcast focused on technologies that support people with disabilities and access needs. Produced by Easterseals Crossroads and the INDATA Project, ATU is widely respected in the assistive technology community for its practical, user-focused discussions and real-world evaluations of tools that make a difference.
Being featured on ATU places Caption Companion within a broader national conversation around accessibility, independence, and inclusive design.
Early in the episode, listeners were introduced to the people behind the product.
Margaret founded Premier Visual Voice in 2012 and has spent her career delivering communication access services, including live captioning, CART, and related accessibility solutions. As someone who is legally blind, Margaret brings lived experience to the design and evaluation of assistive technology. Her work has always focused on reducing barriers—not adding complexity.
Mathew’s background includes sociology, Spanish/English interpreting, American Sign Language interpreting, and more than a decade as a certified captioner. His youngest sister was born deaf, and accessibility has been a personal and professional focus throughout his career. Prior to joining Premier Visual Voice, Mathew led CART and captioning services for a large accessibility provider, giving him deep insight into what works—and what fails—in real-world captioning environments.
Their combined experience shaped the philosophy behind Caption Companion: accessibility should be reliable, simple, and user-controlled.
One of the central questions of the episode was why Caption Companion exists at all, especially in a world full of speech-to-text apps.
Mathew explained that Caption Companion grew out of years of frustration with remote captioning setups in classrooms and meeting spaces. Relying on a user’s laptop microphone, unstable internet connections, and inconsistent software resulted in poor audio—and poor captions.
Even when external microphones were shipped to users, setup issues, pairing problems, and unreliable connections remained. The conclusion was unavoidable:
As long as captioning depended on the user’s hardware, problems would continue.
The solution was to create a dedicated, all-in-one captioning device—preconfigured, audio-optimized, and simple to use. That idea eventually became Caption Companion.
Caption Companion is an 8-inch tablet-based captioning device designed specifically for real-time speech-to-text.
Key components discussed in the episode include:
Unlike general consumer devices, Caption Companion is purpose-built for captioning and communication access.
This was one of the most emphasized points in the conversation.
No. Caption Companion works completely offline.
All speech recognition happens directly on the device. No internet connection, no cloud processing, and no external accounts are required. This allows Caption Companion to function in environments where internet access is unavailable or unreliable, including:
Offline operation also directly supports privacy and data security.
Because Caption Companion processes speech entirely on the device, no audio or text data leaves the unit. This eliminates many of the concerns associated with cloud-based captioning tools.
During the podcast, Mathew explained that many common compliance frameworks—such as HIPAA, FERPA, and COPPA—exist to regulate how data is transmitted and stored. With Caption Companion, there is no external data handling at all.
For users in healthcare, education, legal, or confidential meetings, this local-only approach provides peace of mind and simplifies deployment.
Accuracy was a major topic of discussion, particularly in relation to Caption Companion’s recent certification.
Caption Companion became the first and only automatic speech recognition (ASR) product to pass the NCSP (Speech-to-Text Provider Certification) exam administered by the Global Alliance for Speech-to-Text Captioning.
Margaret shared that Caption Companion achieved a 98.56% score on the exam—an achievement previously limited to human captioners.
Many speech-to-text products advertise high accuracy based on Word Error Rate (WER), which only measures word-for-word correctness. The NCSP exam uses a more rigorous evaluation model that considers:
Errors that change meaning (such as “can” vs. “can’t”) are weighted heavily. This makes the certification a far more realistic measure of caption quality in real-world use.
The successful certification validates Caption Companion as a serious accessibility tool—not just a convenience app.
Both Margaret and Mathew emphasized that caption accuracy is ultimately driven by audio quality.
Even the most advanced speech recognition engine will produce poor results if the audio input is inconsistent, distant, or noisy. Caption Companion addresses this by prioritizing audio capture through:
This audio-first design was repeatedly cited as a key reason Caption Companion performs so well compared to phone-based apps.
In addition to the core device, the episode covered optional accessories that expand how Caption Companion can be used:
The flexibility allows Caption Companion to support everything from one-on-one conversations to large public events.
Speaker identification was addressed candidly.
While some platforms attempt speaker labeling, Mathew explained that accurate speaker diarization is still an unsolved challenge in single-audio-feed environments. Unlike video meetings—where each participant has their own microphone—real-world rooms present a far more complex problem.
Rather than offering unreliable speaker labels, Caption Companion prioritizes accurate, readable captions, which most users prefer over incorrect attribution.
The episode referenced research from the Committee for Communication Access in America (CCAA), which found that if only one communication technology were available in public venues, captions (CART) were the top choice among respondents.
Captions:
Caption Companion supports both open captioning (for everyone) and individual use, making it adaptable to many accessibility strategies.
Margaret shared stories of users whose frustration with phone apps disappeared once they began using Caption Companion. In one example, a grandmother who struggled with interruptions, poor audio, and notifications on her phone regained confidence and independence in everyday conversations.
These stories highlight what the technology is ultimately about: human connection.
Caption Companion’s appearance on Assistive Technology Update reflects growing recognition within the assistive technology community. The full episode (ATU735) is available through Easterseals Crossroads for those who want to listen or read the complete transcript.
Caption Companion is available directly from Premier Visual Voice, with optional accessories and volume purchasing options for organizations and venues.
👉 Visit the Caption Companion product page to explore features, specifications, and ordering details.
