RealComm Global LLC is privately held, founder-led, and based in Florida. This page describes what it is and how it is structured.
A voice technology research company working on multilingual speech models and next-generation communication systems.
It is not a software vendor and does not operate a subscription product. It tried to commercialize one for years: the conferencing platform the company was founded to build was sunset in 2024.
What that leaves is deliberately narrow. A defined research program. A granted United States patent, held by the founder and licensed exclusively to the company. A body of unpublished work held in confidence. Since 2025 the principal build has been the RCG Voice Model. The company is unencumbered and under no obligation to ship anything before it is right.
Founder · Chief Executive Officer · Inventor
Mary Lee Weir founded RealComm Global in 2016 and has directed its research program ever since. She is the named inventor and personal owner of U.S. Patent No. 11,587,561 B2, which she licenses exclusively to the company.
She designed the system and led the full product lifecycle, taking it from an early WebRTC build to a real-time multilingual communications platform that carried translation and sentiment analysis from its first release. Contracted engineers supplied build capacity, and researchers supplied the earliest language and sentiment resources.
In 2017 the platform was demonstrated at the Massachusetts Institute of Technology and at the Startup World Cup, on the strength of its innovation in real-time translation. Development of both the platform and the product continued from there, alongside strategy, intellectual property and research, and commercialization efforts in the United States and internationally.
The emotion-extraction application was filed in October 2019. It covered engineering that was already running in the product, and it followed her conclusion that the pipeline everyone else was refining discarded the part of a message that most often decided its outcome. She extended that stack into AI-driven speech-to-speech translation in 2020.
In 2022 she submitted the voice-driven holographic work to the National Science Foundation as a research proposal. It was not selected for funding. She has financed the majority of the research herself, and she continues to direct it.
Correspondence is read by the founder.
