Valfee (valuable feedback)

Valfee is a gamified educational technology application focused on helping students globally become more effective verbal communicators through optimized feedback, and gamified learning to encourage consistent practice.

On the Valfee platform, students will respond verbally to prompts that have been created based on deep research on the topics students are most excited by, and the topics that will be most important for students to respond verbally to in their academic and professional journeys. When students respond to prompts, the AI engine will produce quantitative scores based on the speed of their speech, the number of filler words, and other indicators that constitute effective speech. These indicators will be extracted from deep audio and semantic analysis of effective speakers in different settings such as presentations, pitches, and podcast interviews. As students achieve higher scores, they will receive coins and in-app incentives, unlock more interesting prompts, and climb the global verbal communication leaderboard.

Up to date, I have conducted deep research to understand what parameters make for effective communication, what topics students most enjoy talking about and communicating in, and what feedback is optimal for the different types of verbal communication struggles. 

77% of U.S. the population has some level of anxiety regarding public speaking and a 2017 BBC piece reported that “oral communication” and “presentation skills” were the No. 1 and No. 4 skills employers seek in applicants, respectively.

Unfortunately, verbal communication education is largely inaccessible. 

Unlike other subjects in education like math or science, there is not a standard verbal communication curriculum. Consequently, people who are not good believe it is a purely “natural skill” that they cannot attain even at the most basic level. The few people that end up being good enough to teach or coach public speaking have to charge inaccessible prices because there isn’t overwhelming demand because most students either believe that they have no hope because they don’t have the natural skills, and more importantly, there is not a rigorous mapping of public speaking skills to other indicators of academic success as these skills are not evaluated in school. This is why while nearly everyone will need verbal communication skills, and very few will use subjects such as biology or precalculus after schooling, the market for science or math tutoring is so much bigger than the market for verbal communication tutoring because students will be evaluated on these things in school and parents want apparent educational outcomes for their children.

The result of this is many smart students who cannot verbally express their intelligence and emotions, and ultimately their success in life is capped. This problem heavily hinders students’ success potential globally.

The solution serves late middle school to high school students globally to reach them early and empower them with the fundamental skills of effective verbal communication, in a gamified and engaging way, to encourage them to consistently practice. These are highly motivated students across all geographies that want to be better verbal communicators, and while they may not specifically be interested or suited for being on their school’s debate team, they want to be future lawyers, politicians, and entrepreneurs, and they want to speak like them. Besides from this debate team, they struggle to find avenues to practice their verbal communication, and receive feedback. 

Valfee will address their need by motivating consistent and long-term practice, and giving them the optimal feedback they need based on their key improvement areas. Unlike other AI speech platforms, I have done a lot of work and thinking about how to gamify the process of learning to speak better to actually make it fun and engaging for every student on the platform. I focus on providing hundreds of questions in many topics so any student can practice talking about the things they love. Additionally, I will use in-app incentives such as coins and leaderboards to encourage additional consistent engagement. I also have done a lot of work to optimize the feedback given for different verbal communication challenges that I will go into in the next section.

With the help of Learn with Leaders, who is also associated with MIT Solve, I have had the opportunity to do a lot of research with students to understand their challenges around verbal communication, and what type of feedback is optimal. With Learn with Leaders, I had the opportunity to work with about 8 students over 2 months and I posted prompts on a website, they responded verbally, and I replied with audio with feedback mechanisms I had devised previously. This allowed me to see if this type of instruction could be effective, and students overwhelmingly felt way more confidence, and their speaking increased significantly throughout the program. After this, I took it a step further and I paid 3 MIT students to teach classes (30 total students) where these students would speak, and the students would give the feedback we have trained them to give based on our research regarding general struggle areas and effective feedback. The students improved significantly, enjoyed the process, and we have a lot of very useful data.

Additionally, to make sure the application is fully gamified and the topics are exciting for students, I launched a global data science competition with 30 global high school teams, to understand what topics are "Trending".

Finally, with the help of Learn with Leaders interns, I have built some code in python with prompts, basic speech scoring, and gamified feedback that I have tested internally, and we are now putting it in front of students to test.

• The product will be called Valfee (valuable feedback), and Learn with Leaders has primarily invested in helping me conduct research by connecting me with students, connecting me with interns, and website design, and incorporating the company Valfee that will be eventually be deploying the product.

• The community I have tested the concept on is quite global. Students who have been exposed to this research and potential solution are from India, Mongolia, Brazil, Australia, and the United States and many more countries. 

• The Trending Topics challenge gave me an extensive dataset of all the type of topics students enjoy verbally communicating in and this data had a similar geographic distribution to the communities I have tested in. 

• The coding of our beta has recently been completed, and it will soon be deployed to at least 30 students in a diversity of geographies to assess if gamified learning of speech is actually engaging to students, and if they see improvement.