April 2022 Summaries

The company traveled to Las Vegas for NAB Show, a major tech show and exhibition in the media and entertainment industry, where they showcased their Autonomous Speech Recognition (ASR) engine. The ASR engine is crucial for greater accessibility in media, as it can save time and money for captioning and other use cases such as digital asset management and media monitoring. The company highlighted how their ASR is vital for live captioning during news broadcasts to ensure accuracy and speed, reducing the chances of error and providing consistent and accurate captions. With high levels of accuracy in 33 languages, the ASR engine has a significant impact on the future of storytelling, enabling media businesses to add captions to content they previously may not have due to time and money constraints. The company believes that automatic speech-to-text at the core of most media workflows is becoming vital for companies to comply with regulations and provide accessible content, as the demand for captioning is increasing, especially among young people who prefer having subtitles on their content. The future of storytelling is an accessible one, and Speechmatics' ASR engine is playing a pivotal part in this journey.

Our latest product release brings significant improvements to our world-leading speech-to-text engine, including Entity Formatting, which has been a highly requested feature and offers improved formatting of entities such as numbers, currencies, percentages, addresses, dates, and times for 11 languages. The new feature uses Inverse Text Normalization (ITN) and provides clear readability in financial documents, saving time and confusion across various use cases like medicine, education, broadcasting, and more. Additionally, we've made improvements to other features such as Speaker Diarization and Flexible Endpointing, and added two new languages - Cantonese and Indonesian, bringing our total number of languages to 33. These updates aim to understand every voice and improve accuracy across all available languages.

The latest update to Speechmatics' Autonomous Speech Recognition software introduces Entity Formatting functionality, which consistently formats entities such as numbers, currencies, and addresses in transcripts, making them more readable and reducing the need for human editing. This addition tackles a challenging task in machine learning known as Inverse Text Normalization and is particularly valuable for enterprise-level customers in industries like media, financial services, and healthcare where accurate formatting of numbers is critical. The update follows Speechmatics' previous advances in Autonomous Speech Recognition technology, which trains on large amounts of unlabeled data without human intervention to deliver comprehensive understanding and reduce AI bias and errors.

This is Earth Day, a day to reflect on the climate crisis and what we can do as individuals and as a collective to mitigate its effects. As a fast-growth voice technology business, Speechmatics acknowledges its own environmental footprint and the impact of its technology on greenhouse gas emissions and electronic waste production. The company has taken steps to reduce its waste, including recycling 50% of its collected waste and offsetting some of its travel impacts through carbon offsetting initiatives. However, it recognizes that more can be done and is committed to making a positive impact on the planet. By embracing sustainability and investing in environmentally-friendly technologies, businesses like Speechmatics can help drive prosperity while protecting the future of our planet.