Listen. Learning a language with AI

It is estimated that 1 billion people are currently learning a language. However, research shows students struggle more with speaking and listening. Better pronunciation means clearer communication, but it's a common challenge for language learners. 

A major reason why pronunciation is so challenging is the exposure to native-like environments. Most language learning relies on generic audio recordings or classroom instruction from teachers. However, these traditional methods fail to give learners tailored feedback based on their specific goals, such as mastering a specific accent or vocabulary. It's extremely difficult to develop natural and fluent pronunciation without a customized experience. 

Furthermore, many language learners are afraid of being judged when practicing in public. The consequences of poor pronunciation skills extend beyond the classroom, impacting personal growth, professional advancement, and cross-cultural understanding. Overcoming this fear is essential for learners to practice their language skills with confidence.

Our solution is a language pronunciation training platform powered by voice cloning technology. It provides learners with an individualized experience to practice speaking in a safe and judged-free environment. 

Here's how this tool works: Students provide a voice sample. Then the AI creates a clone of their voice based on their settings, like accent. Using phonetic analysis, the system checks if the student's pronunciation is accurate. Students can compare their attempts side-by-side with the cloned audio to identify and fix any errors.

While students can upload their own files, the system can also provide specific feedback, such as "You're speaking too fast!" or "Check your entonation."

The key technology behind is voice cloning, which can transfer the pronunciation, accent, and speech patterns from one voice to generate a virtual voice model of another speaker

By leveraging voice cloning AI, our platform eliminates the "listen and repeat" guesswork. Learners get direct, personalized training for their needs. All at home and anytime!

Our emphasis is on non-English speakers who are currently learning a language, focusing on English learners. 

To begin with, we can apply  Listen AI technology regardless of source language, as it is flexible with any speaker. By analyzing a student’s speech patterns, Listen AI creates a cloned voice that matches individual word pronunciations.

Moreover, we plan to connect with local communities through our international team. Oftentimes, English proficiency is one of the most common requirements for job access and career growth. In this way, by teaching English they become capable of thriving within their immediate environments and participating actively in development initiatives in their societies.

Currently, our team is Elly Yu and David Alba. 

Both of us come from the diverse immigrant community. This understanding is deeply rooted in our personal experiences dealing with language learning.

We know the culture and educational systems within our communities. It takes this kind of knowledge to make Listen AI functional with the people that it serves. We intend to make sure that it becomes a user-friendly platform even for those individuals who are not tech-savvy enough.

In everything we do as a team, we take feedback. For instance, we always search for ideas from our communities at all stages. That's how we can develop a product that meets their needs.

We did an MVP in telegram and performed a test for product validation. This was very well received as people loved it.

Based on this, we are now developing the platform and slowly inviting students to participate in user testing. A pilot in three languages – English, Chinese and Spanish will be done first on Listenai.tech within the next 3 months. Now we are setting up servers and databases.

We have a prototype, a business model (open to feedback), and a roadmap of where we are headed. However, we need assistance with finding investors and legal support. 

Unfortunately, we are not well-versed in networking or aware of where we should apply as a company. Our primary concerns are copyright and software rights.

Our solution is unique because it integrates AI voice cloning technology into language learning. In this regard, we redefine traditional teaching by enabling students to listen to their own voices pronouncing “perfectly”. Consequently, our invention may be a stimulus for many students. It provides a tangible and "audible" goal.

To grow both personally and professionally, it is important to be proficient in English.

When an immigrant moves to a country with English as its official language, learning it enables him/her to communicate effectively as well as integrate into the society smoothly.

Also, irrespective of their educational background, people who can speak English in their countries are more likely to get highly paying jobs such as customer service or interpreting which often pay higher than the normal market rates.

Hence, we have created a tool that allows users to quickly improve their spoken language skills whenever they wish and receive individualized feedback. Therefore, we provide flexibility to busy students. They can grow, independently where they are.

Quality of Education (GOAL 4 UN)

We provide students with access to the latest language-learning technology. Our system analyzes pitch patterns, such as speech and intonation, to offer high-quality feedback beyond just phoneme usage.

We use two technologies: ElevenLabs and Praat. ElevenLabs is a voice cloning technology used for creative purposes such as audiobooks or podcasts. However, we want to repurpose it for our needs. 

Praat is software designed for phonetic analysis. We have developed an algorithm that analyzes speech patterns such as intonation or rhythm, providing additional and personalized feedback.

Two people working part-time. We aim to become full-time staff soon, as we launch our pilot and start generating revenue. 

It is been a year, including software development, UI/UX design and research. We are currently on our final stage of development: setting up the database and platform.

Key customers: Students learning English, Spanish or Chinese. NOTE: This is only for our pilot, while we get data and feedback. The tech is available in +20 languages. 

Value: AI platform for language learning, allowing students to hear themselves in their target language + intuitive and inclusive for neurodivergent students (currently investigating about bionic reading and other tech related)

As students, we are looking to save as much as possible: 

1. We are using the GitHub student developer pack to get free credit on database (Google cloud), hosting and domain.  

2. We are using Firebase as hosting thanks to their flexible and "pay as you go" billing plan. 

3. We are applying to ElevenLabs Grants to get free credits and access for three months. 

Thanks to the resources available, our pilot will be low-cost. We plan to generate revenue for three months to establish the company and ensure growth for the future.