Education

The First AI Voice Coach Who Will Actually Make You Work Out Harder

Using AI coaches with the latest technology in voice cloning makes it possible to actually create a user experience that is as close to a face-to-face workout with a personal trainer as it gets, while still being able to serve thousands of users at the same time.

Matt

In 2020, a total of 71,000* Health & Fitness apps launched on the App and Play stores. It’s safe to say that there’s a huge market for physical and mental health.


Market leaders such as Headspace or Peloton are leading the pack (no pun intended) with huge investments of more than $2B** and 70 Million*** users together.


However, the opportunity deriving from this boom in the market of digital fitness products comes with challenges: How can your app succeed when there are so many other options? How can you be more relevant and engaging to your users?

Peloton offers live classes in their app.



How to engage with users in Health & Fitness


Being activity driven, voice and sound are critical user interfaces for Health & Fitness apps. A typical challenge Product Managers face is the lack of opportunities to create personalized and dynamic audio content providing a human-like relationship, while still being scalable.


At the same time, users are becoming more demanding every day. The bar is set high: they expect a variety of running sessions or workouts that fit their different daily moods and routines.


Although voice is becoming a pivotal element for Health & Fitness apps, the state of the art is rather limited: While giving users access to the information they require, nobody loves to run or work out with a GPS-like, robotic voice for guidance and motivation.


What is the optimal user experience?


A good way to approach this is thinking backwards, from the user experience to the product and the tech required to develop it. Let’s think of the most personal workout experience - a personal trainer - and see if we can move that experience into an app.


A personal trainer knows many things about you. They will know who you are, they care for you as a person.They have a close eye on your performance data, not only while you’re working out but also compared to your last sessions. At the same time they will push you. And they will do it A LOT!


So what data do we have today?

  • We likely have personal information such as name, age, goals, etc.
  • We can track and store workout performance data in the backend
  • We can calculate performance during the workout and compare that to legacy data


This is what most apps do already. The distinctive factor to a personal trainer experience is how we deliver that data, or more precisely: How do we connect with the user emotionally?


Two approaches


Live class approaches such as Peloton try to do that with a human trainer, who is able to do one class for many people at once. It’s as close to being in an actual studio class as possible, however it’s not a one-to-one experience. On the other side of things you’ll find many individual-training-based apps, such as running, crossfit and HIIT. They collect tons of user data, meaning they can create a highly personalized experience but they are missing an engaging way to deliver it, at least off-screen.

Individual-training based apps like Adidas Running are already using voice coaches but they lack a human touch and motivation.


The AI Voice Coach


The only way to be able to do that, is text to speech. However, while being able to give users the information, they can’t pull off the motivational value of a human coach.


This is where a new generation of cloned voices come in. They offer all of the advantages and scalability of text to speech, but they are also able to speak to the user like a coach would. Purposely built, a workout voice can hit the same tone of voice and emotion a personal trainer would use, closing the gap between highly personalized workouts and scalability.


Using an API, apps can simply transform their text into human-like speech, with the additional benefit of personalisation, based on real time data: user's name, calories burned, rank in the leaderboard, and even motivational prompts based on your pre-set goals to keep you going. In addition to that, celebrities and influencers can help apps to stand out with the added value of their brand - and their voices.


This is how text to speech used to sound like:


This one has been produced with background music and mastered using api.audio. This makes it sound better, but it still feels pretty lifeless:


And this is how an AI audio coach can sound now, including the full stack of background track and audio mastering:


Conclusion


Using AI coaches with the latest technology in voice cloning, makes it possible to actually create a user experience that is as close to a face-to-face workout with a personal trainer as it gets, while still being able to serve thousands of users at the same time.


Use-case optimised integration helps with easy access and making sure the Product Managers get the job done, fast. If done right, users will feel almost as cared for and challenged, as if they would be working out with a personal trainer. All of that at a fraction of the cost.


About:

Aflorithmic is a London/Barcelona-based technology company. Its api.audio platform enables fully automated, scalable audio production by using synthetic media, voice cloning, and audio mastering, to then deliver it on any device, such as websites, mobile apps, or smart speakers.

With this Audio-As-A-Service, anybody can create beautiful sounding audio, starting from a simple text to including music and complex audio engineering without any previous experience required.

The team consists of highly skilled specialists in machine learning, software development, voice synthesizing, AI research, audio engineering, and product development.


* App Annie’s State Of Mobile 2021 report
** www.crunchbase.com
*** www.businessofapps.com