- New approach increases overall accuracy by 170% and reduces processing time by a minimum of 5X while virtually eliminating most hard costs as it scales to thousands of stations
- Reconfirms guidance for full national launch in second half of 2021
BOULDER, CO / ACCESSWIRE / June 22, 2021 / Auddia Inc. (NASDAQ:AUUD) (NASDAQ:AUUDW) ("Auddia" or the "Company"), developer of a proprietary AI platform for audio and innovative technologies for podcasts that is reinventing how consumers engage with audio, today announced a major advancement in its proprietary technology at the core of its Artificial Intelligence engine. By leveraging precise audio and metadata from radio stations, Auddia will reduce the costs of processing audio content for AI training and validation to near zero and concurrently realize vastly superior improvements in accuracy. The new AI processing methodology gives the company near real-time data processing capabilities which is expected to improve overall performance of the Auddia platform and reduce the onboarding time for stations by a factor of five as the company scales to thousands of stations.
Auddia expects to use the new AI methodology for its trials with Lakes Media and Sonoma Media that were announced previously and will begin shortly after the July 4th holiday. The Company continues to expect to report consumer interest and subscription pricing from both Lakes Media and Sonoma Media audiences in anticipation of its full national launch in the second half of 2021.
Peter Shoebridge, Auddia's Chief Technology Officer, explained, "Our latest advancement in AI takes advantage of what we always understood to be one of the most valuable elements of the audio content ecosystem, which is the abundance and open availability of audio data. Accurately tagging that audio data with precise metadata is the ultimate objective, and our new methodology enables us to meet that objective. Recent test results that allow us to compare our new approach to previous methods reveals orders of magnitude improvement in areas that are critical to the business, including accuracy, speed and timeliness of AI training, and the costs of operation."
Auddia Chief Executive Officer Michael Lawless commented, "While our initial AI engine produced suitable results, this major advancement in technical capability is a game changer for the company. We always had a sense that a significant leap forward was on the horizon, so we are pleased to achieve this major milestone now, as we launch our first commercial consumer facing trials and anticipate a rapid increase in radio station deployments. The expectation is that the user experience will be positively impacted as a result of this advancement."
The latest technology advancement allows Auddia to train its AI engine with significantly more data, in less time and at a greatly reduced cost. The new approach generates far greater and quicker content identification accuracy. For example, to train the AI model on a group of six radio stations, the previous approach required a minimum of 50 hours of audio to achieve satisfactory results. This human-intensive process would take five days with hard costs over $2,100. With the latest advancement, the same group of stations can be trained on 1008 hours of audio data - a 20X increase in data volume - with zero hard costs, completed in a single day (versus 5 days) and yield far greater accuracy. When developing AI, training with more data is always preferred, but to process 1008 hours of radio audio using the previous approach would have resulted in hard costs over $40,000.
Working with significantly more audio data that comes with precise metadata, results in highly accurate AI algorithms. At the time of this announcement the improvement in content identification accuracy is higher than 170%, with potential to improve further. These algorithms are expected to produce greater accuracy in all applications, including when used against radio stations that have had no AI training whatsoever.
About Auddia Inc.
Auddia is reinventing how consumers engage with audio through the development of a proprietary AI platform for audio and innovative technologies for podcasts. Auddia offers two industry firsts -- the ability to listen to any AM/FM radio station with added personalized content and no commercials as well as podcasts with an interactive digital feed that supports deeper stories and delivers digital revenue to podcasters. Both offerings address large and rapidly growing audiences with strong purchase intent. For more information, visit: www.auddia.com
This press release contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933 and Section 21E of the Securities Exchange Act of 1934 about the Company's current expectations about future results, performance, prospects and opportunities, including, without limitation, statements regarding the anticipated use of proceeds from the offering. Statements that are not historical facts, such as "anticipates," "believes" and "expects" or similar expressions, are forward-looking statements. These forward-looking statements are based on the current plans and expectations of management and are subject to a number of uncertainties and risks that could significantly affect the Company's current plans and expectations, as well as future results of operations and financial condition. These and other risks and uncertainties are discussed more fully in our filings with the Securities and Exchange Commission. Readers are encouraged to review the section titled "Risk Factors" in the Company's Annual Report on Form 10-K for the year ended December 31, 2020, as well as other disclosures contained in the Annual Report and subsequent filings made with the Securities and Exchange Commission. Forward-looking statements contained in this announcement are made as of this date and the Company undertakes no obligation to publicly update or revise any forward-looking statements, whether as a result of new information, future events or otherwise.
SOURCE: Auddia Inc.
View source version on accesswire.com: