Nyansapo: Supplenting Teachers Digitally

The United Nations Educational, Scientific and Cultural Organization (UNESCO) estimates that globally, at least 773 million youth and adults still cannot read and write, and 250 million children are failing to acquire basic literacy skills. The lack of foundational skills excludes youth and adults from the opportunity of participating and contributing to their communities and societies, which greatly impacts the social and economic development of these countries. Sub-Saharan Africa is one of the most affected areas in the world, with there being over 32 million children of primary school age, 54% of which are girls, who remain uneducated. With more than half of the children in Saharan Africa receiving education for less than four years, it is clear there needs to be a drive to encourage secondary education.

The typical Kenyan classroom is overcrowded with about 48 or more students of various ages and maturity levels within the same grade. Due to the lack of resources and inability to effectively teach such a wide variety of students, many students will begin to fall behind, moving through the educational system with great fundamental misunderstandings, causing their learning to reach a standstill. Being unable to pass placement tests for secondary education discourages the students and leaves them in a position where they are forced to leave school early. With little to no education, these children are thrown into society, yet must provide for their family, and are left with low paying jobs that allow them to barely make ends meet, furthering the continuation of the cycle of poverty.

According to (Muralidharan, et al., 2019) One of the leading explanations for this low productivity and learning of children is that existing patterns of education spending and instruction may not alleviate a key binding constraint to learning, which is the mismatch between the level of classroom instruction and student learning levels. And because of this most children are unable to benefit from traditional classroom instructions.  Teaching at the Right Level (TaRL) is one of the teachings and learning method that have been developed to address this issue of learning. In TaRL students are regularly access to understand their current needs and instructions are tailored specifically to the current needs of the children.

  There is the need for the use of technology to help scale up this wonderful method that addresses the issue of child education at the core. Nyansapo AI aims to develop solutions that can assist instructors to assess children and provide children with specific instructions that the children need. 

Nyansapo is an app created to streamline the teaching process of using Teaching at the Right Level. It is built not to replace, but rather assist the teacher in the time-consuming areas of assessing, grading, and grouping students. This can turn the process of assessing, grading, and grouping students from a whole school day to just an hour or two.

This is made possible by Nyansapo’s speech to text model, trained with data across Kenya to capture all speakers at a 95% accuracy rate. Because of this, students can simply use a device, whether a tablet or phone, to listen to their responses to reading prompts being shown on the app instead of the teacher having to test every student individually. This will greatly speed up the assessment process and grading process as the student's results will automatically be recorded. In addition, Nyansapo will automatically sort students into groups of similar skill levels.

The app currently can give assessments for both literacy and numeracy. In addition, it is built in a way that can be applied to tackle any problem that is using the teaching at the right level methodology, and the tests for literacy and numeracy can be customized to a customer's liking. 

Our end goal is improving the lives of students by making sure they have the core skills that are necessary to pass placement tests in order to advance their education. Children are currently underserved in Kenya as a result of overcrowded classrooms. This is especially difficult in the public schools with students from low-income households. They are the ones primarily supported by the NGO catch up programs using Teaching at the Right Level that we partner with. Our solution will help those catch up programs become more effective through freeing up the teachers time. We do this through improving the tools available to the teacher that will take care of assessing, grading, and grouping so the teacher can put all of their effort on teaching the students and therefore making the biggest possible impact.  

The idea of Nyansapo was born when our CEO Mumbe, a native Kenyan and NGO worker very aware of literacy catchup programs and their shortcomings got in contact with Edward, a Penn State Student from Ghana. Since then, it has expanded across Penn State with currently 5 other Penn State students working on it. In addition, it has brought more people in from Kenya with the primary creator of the app, Justice, being a native Kenyan. Mumbe, being raised in Kenya, is very aware of how to move our mission forward in Kenya. In addition, those furthest from it the 4 of the PSU students, will be spending 4 weeks in Kenya working to both advance the venture and also gain first hand experience of what they have read so much about. 

We are moving beyond our prototype now. Though students have done well at creating the initial versions of the app we now have to find a way to expand across many more platforms and operating systems with a more professional app. For this reason we hope to form partnerships with those that can develop and launch our app at an affordable price.

One of the main innovations of Nyansapo AI is that we took an amazingly effective technology, which is speech recognition and applied it in the context of education. This innovation will allow rapid assessment of children to enable swift delivery of targeted instruction to children all across the world. The biggest innovation for me as a co-founder is that Nyansapo AI focuses on the less privileged children in the school. The children that have been left behind are unable to learn in school. Nyansapo AI aims to use technolgy together with people to bring these children back into the wonderful world of science and technology. If this is not a great innovation, then I do not know what an innovation should be.

Instead of replacing a teacher with technology that simply isn’t as effective as having a real instructor we streamline the teachers process of interacting with the students. point is that if it reduces the testing from 1 day to 1-2 hours in a 10 day camp, that effectively grants almost 10% more time at minimal cost to the program.

In addition, one of Nyansapo’s main selling points is that it allows customers to customize their programs and assessments as they see fit. This could come in the form of their own questions and format for assessments in core focused areas like literacy, or even expand out beyond literacy and numeracy 

Within the next year we hope to make our first sales to at least 5 major NGOs for the app to be used in bootcamps no longer under our direct supervision. From these partnerships we hoping the app will reach three thousand students and help them with their literacy advancement. This will both form a foundation for our business revenue and allow us to see and adapt to any changes. From a technical perspective we hope to create an offline speech to text model as due to budget and other constraints our current one requires access to the internet.

After the year has passed and we have laid the foundation of Nyanapo we hope to expand all across the continent. First in keeping with where the members roots are we will expand to Ghana and then after that across all of Africa and hopefully become a standard part of every Teaching at the Right Level based bootcamp. As a standard part of the now much more efficient and cost effective bootcamps we hope to then advocate for not just Nyansapo but also all of TARL and the bootcamps. Finally we will also customize the app as requested to go beyond just literacy testing with the TARL method and make our app useful for all NGOs, schools, and businesses. 

Our main measurement revolves around how much time is saved by using Nyansapo and therefore the increase of contact hours between teacher and student. This May we will be conducting an experiment that will measure the amount of time saved using Nyansapo based off the number of students, teachers, and tablets. With those figures we will create a model that will be able to estimate how much time is being saved in camps using it, based on the anonymized data we collect.

In addition, we will measure impact by examining what communities the Nyansapo app is being used in. Though we want it across Kenya the emphasis must be on aiding low income and marginalized communities. 

Currently 18.5% of Kenyans cannot read or write a simple statement.  This can be attributed to overcrowded classrooms and underfunded schools, which lead to children falling behind in class.  Supplemental programs are important for helping students catch up but they cannot reach all the students that are interested.  Additionally, there are a lot of errors and delays in getting the program data to the necessary people. 

            If there is an app that accurately assess and grades students, then teachers will have more time to teach foundational skills.  This increases the contact hours between the teacher and students.  The time saved using the virtual assessments allows supplemental programs to spend more time teaching weaker learners or allow supplemental programs to hold more camps throughout the year to reach more students.  In both of these scenarios more students will have a better understanding of the subject and the literacy rate in Kenya will go up.

            Additionally, if there is an app that collects quality data that can be viewed on multiple deices, then supplemental program directors can quickly receive the data and make decisions.  This will allow directors to tailor their programs to better suit the needs of their students.

            The inputs involved for this solution are the existing teaching at the right level methods and curriculum, programs and schools dedicated to increasing literacy rates, smart devices, and Microsoft Azure speech to text model.  These inputs will allow for creation of an app that can provide TaRL assessments, group students, track progress, recommend activities, assist in scheduling, and rent tablets.  This app will make the assessing and grading portion of the TaRL process quicker.  It will also allow programs to quickly examine results.  This allows programs to reach and significantly impact more students, improving their learning outcomes.

The Nyansapo AI leverages the power of speech recognition to administer literacy assessment to children between the age of 6 and 16. The children are provided with a series of sentences, words, and letters to read. As the children read these sentences, words, and letters the mobile applications convert the audio of children reading into text. This converted text is then evaluated with a custom Natural language processing model to identify words mispronounced by the child during their reading.

A similar method is applied in the numeracy assessment where both speech recognition and computer vision or handwriting recognition, to be specific, are used to evaluate answers provided by children to simple prompts. Children are assigned a learning level based on their performance during the assessment. For literacy assessment there is the Letter, work, paragraph, and story levels, and for numeracy assessment there are number recognition, addition, subtraction, multiplication, and division levels in that order. 

We’re proud to be a Kenyan owned and woman owned company as a result of our CEO Mumbe. The rest of the team finds strength in being a mixture of Kenyan and other African students and professionals alongside students from Penn State University. When the business launches the employment and management will shift entirely to Kenya and then will expand first to Ghana and then hopefully will expand to Ghana followed by all of Africa. 

The value that we provide is saving the teachers time so that they can focus on spending more time teaching the students instead of assessing, grading, and grouping them. This carries over on impacting the students because normally students have 8 days in a standard literacy bootcamp to learn from the teacher as the assessing takes about 2 days. We can decrease it to a day or less. That’s 12% more time that the NGO program gets to have teaching students at an incredibly low cost. There are other reasons the NGOs would want the product such as they can be confident that the testing results are accurate and authentic, as we’ve been told they sometimes question the accuracy of the final tests the teachers report. Additionally, we’ve created a dashboard where they can easily see and analyze the data that they have collected. It also services teachers in another way by providing suggestions for engaging activities to do with the students that we have found have worked best in the past.

            That product being the app is our primary service and its distribution will be fairly simple. It will be available first on Android Play Store for download, then later on the Windows Store and App Store. From there they can create an account and use either the free version of the app or the member or premium accounts. The member accounts will be able to carry out several bootcamps a month. Meanwhile the premium accounts will be able to carry out unlimited testing per month and the member will be able to submit custom tests that we will program into the app and make available to them. We have not yet decided on the monthly costs to each tier as we are not sure how much NGOs are prepared to pay. This is currently the focus of our customer research.

            Our other revenue stream will be from renting out tablets to schools and NGOs that need them. For this service to work properly we will have to determine a reliable and fast delivery service in Kenya to partner with in order to ship our tablets between camp locations. We have not yet determined a price as we have to ensure that it is enough to cover shipping the tablets and getting insurance on tablets along with a bit more to help cover Nyansapo’s ongoing expenses.

As for finding customers we have our website at https://www.nyansapoai.com/, and we hope to find customers initially by harnessing Mumbe’s connections in the largest NGOs that do literacy boot camps. 

Under our current business model, we have two primary revenues of money. The first is the cost of subscriptions for the pro and premium organization accounts of the Nyansapo app. There is a free version that allows for a once a month testing for a small group as well. A second potential revenue stream is that some NGOs or any organization may want to save time using Nyansapo but don’t have the capital to buy their own tablets. We’ll therefore arrange an affordable tablet rental business so that they won’t have to invest in such a purchase.

We have not yet attempted sales because the bootcamps that we have used Nyansapo in have been used for research and testing of our app. However, we have had a good deal of success in pitch competitions by winning the Nittany AI for $10,000 and also the Microsoft Imagine Cup for $2,500