EZSpeech

• 20% of kids age 3-10 have various speech disorders
• worldwide over 100 million kids and adults

Problems: 

1) Shortage of SLPs – currently there are 400-800 kids per one speech pathologist (SLP). Waiting list for SLPs even in developed countries is 3-6 months. 

2) High fees are not affordable by many families – Fees per session ranges between $60 and $120. Insurance covers only limited number of sessions, if covers at all.

3)  Low efficiency of the treatment - Normally SLP spends with a patient~30 minutes twice a week, while effective treatment requires daily practice. Parents rarely make sure that children practice daily

We used VSM (Video Self Modeling) approach, which proved to be more effective than any other methods of treatments, including imitation. A child sees himself\ herself on a PC screen when performing exercises. Animated mask and voice instructions enable a child to see and understand what he/she should be aiming to achieve.

Artificial Intelligence (AI) identifies correct and incorrect performance of the exercises. In the process, the algorithm is capable of understanding and interpreting the exact errors. It identifies precisely which group of muscles should correct their movements and how. This process enables the AI to give precise feedback and instructions to the user on how to correct the errors. 

We use computer vision for both mimic and voice tracking. A low-level model with 90-120 key points involved in facial expressions plus a higher level model that describes the rules for performing exercises

For voice at certain parts of the exercise process, we record the sound and also use the computer vision to analyze its spectrogram to classify the sound - whether the result sounds correctly. For example “Sh" or "S", " R " or " L " and so on. Any sound can be checked separately in any part of the word (beginning, middle or end) 

We offer:

1) Assessment and interpretation

The App can assess and interpret performance of speech therapy exercises. Automation of this process increases reliance on status and progress of the Subject, decrease human error and avoid subjectivity in the process

2) Personalized guided performance

User receives tailored feedback and guidance in real-time, specific to the case  

3) Personalized plan and Progress reports

Set of exercises are tailored to the needs of the specific User (personalization of training): 

At the time of registration a status of the User is determined based on automated "Initial Assessment" and the App generates the "Individual Plan of Care" (IPOC) with a baseline and recommendations.

The App monitors the progress according to IPOC on a regular basis (the process is fully automated) and generates regular reports and recommendations based on the progress achieved, including but not limited to modification of IPOC

4) Animation & Gamification

Animated masks and gamification increase child’s engagement in the learning process as well as increase motivation

5) Low price

Our fees is only $1-$5 per session, I.e. 10 time cheaper then a human SLP with higher accuracy and more fun

Video demo demonstrate all key features of our App (Refer section "More about your solution")

We follow a UN principle "Leave no one behind"

Correct speech is a crucial element of any educational process ensuring ultimately both social and economic integration essential for the well-being of both individual and of the entire society.

Due to high cost of the treatment and the fact that speech therapy is not always covered by insurance, a lot of families simply do not have access to speech therapy.

Our target audience includes:

1) children and adults from low-income families, who can not afford SLP. Our plan to expand in US and worldwide, primarily in Latin America, Asia and Africa

2) children with cerebral palsy and autism - We hope gamification of the process and innovative approach will drastically increase their engagement and make a boring and difficult process a little less challenging for them. 

3) non-English speaking children and adults (in U.S. and worldwide) - in our Center we serve immigrants. Many of them don't speak English and don't always understand explanation of SLP.  Visual demonstration of an exercises as well as step-by-step guidance does not require knowledge of English. So non-English speaking patients can be treated worldwide. 

Each of our key team members are very well familiar with the problem and has personal attachment to it.

Firouza Eshonova (Uzbek)  Firouza's son has suffered from severe speech problems and received a long-lasting treatment in Speech Center. That's why she came up with the idea for this project. He background and experience helped her to organize a team of highly qualified experts and professionals and manage the project effectively. She established relationship with a number of institutions and organizations, including international, to ensure that the product we are developing serves the needs of the most vulnerable groups of population.

Adelina Samuels (US) is a Qualified Speech Language Pathologist with over 20 years of experience. She is an Active member of ASHA (American Speech and Hearing Association). She is a Clinical Director and Founder of Medical Group Little Steps. All her professional career Adelina have been providing serviced to low income patients with speech and language disorders and disabilities, as well as army personnel and their children that had suffered from neurological and cognitive disorders and disabilities. Out of 200 patients she serves within her Speech Center approximately 180 are from low-income families.

Denis Bondarenko (Ukrainian) has nearly 20 years of experience in IT industry. Since 2001 Denis has been engaged in computer vision, 3D technologies, and video data enrichment.  Denis's son has speech disorder so Denis is keen to develop an App to help his son and other children. Over the last year Denis is working along with his son's SLP on a regular basis.   

1) the key goal is to expand our network, especially in developing countries, specifically Latin America, Asia and Africa

2) to access the mentorship and coaching advice from experts and professionals on how to improve our App, as well as how we can adapt our technology to other areas

2) to raise awareness about our solution so schools and communities can use our App to cover their specific needs

Based on our patents' search there are no solutions on the market similar to ours. We are finalizing our application for IP registration and plan to submit it shortly

Speech therapy services are represented by off-line and on-line sessions (by a human SLP ). There are various Apps on the market which demonstrate speech exercise. But none of them able to access the performance. Our solution offers a virtual SLP (on-line simulator), which enables kids to practice independently without supervision of a SLP or a parent. 

Similar technology is used in apps for yoga, fitness or back exercises (for example https://www.kaiahealth.com/  ). But no one has developed a software for speech therapy, especially multimodal.

In terms of innovations, we have several:

1) Assessment and interpretation

Similar technology exists, but not for speech therapy which involves face and sounds. We have developed algorithms specifically for the purpose of face muscles, gestures and voice.

Personalized feedback which identifies an error and tells the User how to correct it (in real time) is not available in any other speech-therapy Apps

2) Multimodality - to the best of our knowledge the existent software covers maximum two modules, where is our will combine three: face, gestures and voice

3) Dataset and Expertise - we do not use open source, we generated and expanding our own data-base of specific datasets with domain valuable data

4) Automated Decisions Support System - stats and progress reports generated by the app can be used by SLPs, School's and Clininc's management, as well as Insurance companies for monitoring the progress, modification of IPOC, methodological purposes, management of processes etc  

New approach to the problem:

1) speech therapy will be available for kids from underprivileged families

2) effectiveness of treatment will increase as kids will combine their practice with human SLP with daily exercises through our App

3) hybrid-therapy and Web-simulators - can be used to optimize the work-load on SLPs and operating expenses of Speech Centers and Hospitals

4) objective assessment of progress generated by our App can be submitted directly to SLP or medical records to analyze the effectiveness of treatment  

Access to App is available upon request

We are currently piloting the App in several Rehabilitation Centers

NEXT YEAR – Pilots, partnering and improvements:  

Our impact goal for the next year is to make at least 3-4 thousand kids confident that they can improve their speech disorders, and to equip their parents and SLPs with a tool that assist them in achieving better results within a shorter timeframe.  

We will achieve these goals by: 

• Raising awareness via ASHA and public sources about our APP 

• Partnering with schools, speech centers, rehabs and insurance companies 

• Receiving and responding to feedbacks and recommendations from users (both private and institutional) on how to improve our APP 

Upon completion of successful development stage we will move to growth and scale-up stage 

NEXT FIVE YEARS – Scale-up and increase effectiveness of the treatment  

Our longer-term impact goals are: 

• Change the existent practice when only limited number of kids have access to speech therapy services. Our goal is to make it available to anyone who needs it  

• Substantially increase effectiveness of speech therapy through implementation of hybrid approach to treatment (combining human SLP with a virtual simulator) based on technological innovations (computer vision, AI, machine learning etc) 

Ultimately, this will fundamentally transform the existent practices and approaches to speech therapy treatment. 

SLPs around the world will be able to use the best practices and virtual support available from the App 

Over 40 million children worldwide will get access and benefit one way or another from using our App. 

We will achieve these goals by: 

• Scale up worldwide using direct marketing, local telemedicine and/or on-line education platforms, forming partnerships etc (the exact way will depend on country specifics)   

• Introducing various tariffs to make the App available to broader audience (including a free-tariff funded from proceeds from the adds) 

• Ensure that our APP is HIPAA compliant or compliant with other countries local rules and regulations 

• Develop medical decision support system in the area of speech therapy using machine learning and AI 

1) we have developed and successfully piloted MVP in two Speech Centers. Focus groups in both centers (SLPs, kids and their parents) confirmed their strong interest to use the App

2) throughout the entire process of development we discuss the App design and functions with potential users to receive their feedback. Our focus group include over 110 kids age 3-10 in US, Uzbekistan and Russia 

3) Based on feedback received during two pilots in Speech Centers we have improved certain function to increase users experience, especially tailored for kids with disabilities (for example who are not able to press buttons or scroll the screen)

3) We have agreed pilots in several other locations in US and Russia (may be we will have to reconsider location based on political situation)

4) We have established contacts with several Foundations and Organizations which support schools for kids with disabilities and communities with low-income families

5) We have drafted the detailed plan for the next stage of our APP, which includes VR/AR and found highly skilled IT developers to help us with this module 

IF we make our App available to children with speech disorders as well as to companies and organization treating those children 

THEN there is a much greater possibility that over 40 million kids around a globe get higher chances for inclusivity, equal opportunities and self confidence

BECAUSE We increase accessibility, affordability and efficiency of speech therapy services   

Empowerment of kids with disabilities have positive spill-over for both the well-being of the families and communities and for public expenditure structure, making ultimately economic growth more inclusive and equitable, leaving no one behind 

Core technologies: 

- Computer vision

- AI, DL and ML (executed and piloted) 

- AR/VR (in process of development)

Core technology based on highly generalizable recognition algorithm

Underlying technology: Due to the architecture all scenarios of face
movements are covered and interpreted. ANN identifies, trace and interprets movements of face muscles. We use 90-120 key points to insure high precision and can assess performance even if a users moves head. We have already reached 83% to 90% accuracy and targeting to reach 95%.

Multimodality The App processes combinations of contractions of muscles of the face, the voice, the tongue and lips simultaneously

Transfer learning from SOTA + production approved

It is based on: a) state-of-the-art artificial neural networks b) production of approved neural networks and ensemble architectures
c) transfer learning to the subject area

Specific datasets + scalability by architecture

The architecture is made to scale. i.e. covers new cases within the subject area so that their development is cheaper and faster than from scratch. A dataset is prepared that is specific to the subject area. It has value in itself, because it is not just a video, but marked-up frames with measured reactions of the neural network: expected vs fact

Our goal was to create a team of highly qualified professionals who:

a) understand the problem and the target audience 

b) can execute technological solution and 

c) can execute the project based on principle "impact for profit", i.e. combination of business-for-profit with the main goal to help those who is in need and to achieve the greatest impact 

We didn't specifically search people based on their gender, color or ethnic background. 

Nevertheless, 2 of 3 our leaders of women and both are minorities. Adelina is a refugee and Firouza is Asian. Both Adelina and Firouza are self-made women with outstanding professional careers. Our IT Lead is Ukrainian, so as several of our IT developers. 

So we consider our team quite diverse in all aspects.

We established contact with several foundations that support people with disabilities, including Global Disability Foundation. 

Our goal for the next 12-24 month to develop partnership and alliance with Charities and Foundation. We are discussing a membership in our Board with a couple of their representatives.

OUR VALUE PROPOSITION: Leave no one behind - to make speech therapy available to any child who needs it. Let's give equal opportunities for education for children in need

CUSTOMERS AND BENEFICIARIES: 

1) Companies and organizations: Schools, Speech Centers, Hospitals, Rehabilitation Centers etc, which experience shortage of qualified SLPs and/or can not afford to keep sufficient number of SLP to service their patients. Insurance companies are considered as alliances. 

2) Retail: Low-income families who can not afford speech pathologist or do not have an access to SLP

Ultimate beneficiaries: The entire Society, Federal and Municipal Budgets

PRODUCT WE OFFER 

1) B2B  - license to use the software (SaaS) for in-house purposes and/or distant monitoring of out-patients 

Purposes:

- Increase of number of people served by applying a hybrid treatment and to address shortage of SLPs by combination of human SLP with virtual simulator

- Objective and independent valuation of effectiveness of therapy and progress made

- Decrease of operating expenses

- Increase of effectiveness of the treatment

2) B2C (retail) - subscription to App (Web and/or Mobile) 

Purpose: Home exercises between visits to SLP to practice movements of articulators and pronunciation until these become automated. App can be used as a matching platform for both parents and SLPs who can provide feedback and recommendation based on dashboard and statistics 

WHY DO THEY WHANT OR NEED OUR PRODUCT?

1) Shortage of SLPs 

100 million children worldwide suffers from speech disorders. Currently there are 400-800 kids per one SLP. There is a recognized shortage of SLPs worldwide.

2) High cost of SLPs

The average fees for on-line SLP is $55-$60 ($100-150 for off-line). Insurance covers only 35 sessions (if it covers at all). Most of the low income families do not have even this coverage and can not afford any SLP.

Many Speech Centers do not have sufficient budget for adequate number of SLP.

The fees for our virtual SLP is only $1-$5 per session compared to $55 for online. 

3) Objective stats and progress report

The App will generate stats and progress report 

So far, the project was founded with investments from founders and an angel investor. To cover cost of servers we used our investments and grants from AWS (common practice for startups). Currently we are not planning to attract additional funding.

Based on our forecast (3 year business plan and sensitivity analysis) we will reach a break-even within 5-6 months after launch and going forward expect to be profitable and self-funded.

But our ultimate goal is to reach-out any kid who requires speech therapy. There are millions of children worldwide who can not afford paying even $1-$3 per session and there are many schools, speech and rehab centers which lack sufficient funding to acquire license for our product.

For this segment we consider alternative sources of funding:

1) free App/License with built-in adds (common practice) 

2) raising charity money to acquire licenses for specific School, Speech Center or Rehab  - third parties (people or companies) can to purchase subscription or license on behalf of any child or organization 

So far we have:

1) made investments from founders

2) raised investments from angel

3) received infrastructure grants (to use GPU servers from AWS and Yandex)