The incredibly popular language learning app ELSA Speak enters a new era with powerful new generative AI features. 

Leveraging its state-of-the-art speech recognition technology, ELSA today announces the development of its powerful new capability: ELSA Voice AI Tutor. Powered by proprietary speech recognition technology and informed by one of the world’s largest accented datasets, this new voice-enabled capability will showcase the power of generative AI in language education. 

ELSA has always been at the forefront of the AI revolution in language learning and spoken English proficiency. Since 2016, millions of users have improved their clarity and confidence in English using the app. With over 50 million downloads the company has been able to gather millions of hours of accented spoken English audio data from English learners around the world. 

Meet ELSA Voice AI Tutor – a free near-human-like English Speaking tutor currently in exclusive private beta. 

Many of you might have tried out and experienced the recently published AI Chatbots like OpenAI’s ChatGPT, or Google’s Bard and have been blown away by its intelligence and the responses they generate based on our questions and prompts through texts. 

However, until now, no teaching application has convincingly reproduced a natural and seamless conversation between two language learners speaking to each other. 

That’s about to change. 

Our multidisciplinary team of AI scientists, expert linguists, and teachers are leveraging generative AI technology to develop what we call ELSA Voice AI Tutor to achieve that very goal.

ELSA Voice AI Tutor will act as your very own Conversational AI tutor: One that allows users to speak in free-flowing, spontaneous and engaging real-life conversations on any topic or scenario. 

Learners will then receive tailored feedback across multiple dimensions on their English speaking proficiency. These include pronunciation, vocabulary, grammar, intonation, rhythm and more.  

“For the first time, an AI tutorial experience will be as close to a human tutorial as possible,” says  Vu Van, CEO and co-founder of ELSA Speak. “AI will allow for long conversations with students, and if they don’t understand something, they can stop and ask the AI,  ‘What does that mean? Can you go deeper? Can you give me some examples?’” 

Upcoming powerful features

• Voice-based interaction: Using specific keyword activation, learners can simply say “Hey ELSA” to activate ELSA Voice AI Tutor. Once activated, ELSA’s AI tutor will be there to guide you through your learning journey in the most interactive and immersive experience.
  
• Choose your own adventure: Each English student has different goals. Some want to improve their English at work, others want to make social interaction more enjoyable, and some simply want to order dinner at a restaurant. ELSA Voice AI Tutor will allow them to choose the situations and the topic that match their goals. It will generate conversation and respond naturally in a role-play scenario. Conversations will adapt based on the learner’s English level, personal style and speaking context/environment.
  
• Ask ELSA Anything (AEA): ELSA Voice AI Tutor allows users to interact and follow up with ELSA in real-time for any questions they have on their speaking challenges, just as they would with a teacher in a one-to-one classroom. 
  

Learners can ask ELSA anything that they need help with. For example, they can ask for more detail about the mistakes they made, request extra examples or seek clarification. 

Moreover, feedback is hyper personalized for each learner. It will remember all of their past mistakes and use information from previous practice sessions, getting to know a user’s individual strengths and weaknesses. All in all, it’s like having the benefits of a private tutor, for free. Moreover, it can be delivered at scale which is great for users in a corporate or educational setting.  

Where These New Capabilities Shine 🌟

How will ELSA’s use of generative AI benefit learners in different ways than other apps? Imagine these situations:

Test preparation 🧑‍🏫

One of ELSA Speak’s most powerful features is its ability to present users with adaptive practice questions. We plan on taking these practice questions to a whole new level of interactivity using Generative AI. Preparing for the IELTS Speaking exam, for example, will begin to look more like this interaction:

English for work 👩‍💼

Perhaps the learner wants to employ ELSA to help her with a delicate situation such as asking for a raise. That conversation could go like this:

Casual conversation 🗣️

There is perhaps no better way to learn to speak a new language correctly than through casual, no stress conversations on topics of interest to the learner. Generative AI now makes this possible. We’re planning to make open-ended interactions like this one possible:

Speaking practice for schools 👩‍🏫

Research has shown that up to 60% of ELL students struggle with English speaking, and are not interested in learning English as a result. 

The new Voice AI Tutor feature allows teachers to assign speaking tasks. Students can then practice at home for HOURS with ELSA’s AI Tutor. 

It is infinitely patient at pointing out mistakes and giving tips to improve in real time, and with no added embarrassment of making mistakes in front of a person.

Homework tasks might include: 

• Telling ELSA about your weekend plans 🏖️
• Having a conversation with a new friend (ELSA) and introducing yourself 🗣️
• Practicing the new vocabulary you’ve learned in class and talk to ELSA about your favorite sports 🏂

The Future of ELSA Speak 

For years, ELSA Speak has delighted and empowered learners, teachers and business leaders in their careers and social lives. Our goal has always been to engage and build confidence in our users. 

We are now stepping into a future with more powerful technology. Don’t miss your chance to meet ELSA Voice AI Tutor – a near-human-like English Speaking tutor. 

Interested in learning more? Sign up to get early access.