English Helper’s AI-based solution to develop English language skills is supporting the students of more than 25,000 digitally equipped schools in multiple states. It aims to reach over 1,00,000 schools in 2021-22
For students in government schools in India, learning English can be challenging for multiple reasons, including lack of exposure to the language, teacher shortages, and the language proficiency of educators. Many of these schools, especially in rural areas, use the local vernacular or state’s official language as the medium of instruction. Acquiring English language skills can boost prospects for these learners —from opportunities for further studies to eligibility for suitable jobs leading to improvement of livelihoods and incomes of rural households across the country.
To remove barriers to improving English language skills in government schools across India, EnglishHelper (EH), a language learning technology company and social enterprise, created ReadToMe, an AI software that facilitates multi-sensory reading and comprehension. Intel, in collaboration with Amazon Web Services (AWS), enabled EH to effectively scale this solution to support the students of more than 25,000 digitally equipped schools in multiple states, with the aim to grow to beyond 1,00,000 schools in 2021-22.
“Technology is transforming education, changing how students learn and empowering them at every stage of their journey. The ability to read, write, and speak English can open a multitude of opportunities for children. Intel architecture and AWS cloud enabled scalable computing capacity keep costs low and deliver advanced features that enhance the learning experience under EH’s RightToRead program. Together with EH and Schoolnet, we are working to expand this collaboration to other states in India to elevate learning outcomes across the country,” says Akanksha Bilani, Regional Alliance Head – Asia Pacific & Japan at Intel.
ReadToMe has been implemented for students from grades 1 through 12 under EH and Schoolnet’s RightToRead program. Independent evaluations have found that students undertaking ReadToMe-enabled classes showed 20–40% higher improvement in reading and comprehension. The programme also includes implementation of a learning solution called EnglishBolo to equip teachers with improved proficiency and familiarity with ReadToMe through a blend of self-learning and live tutoring sessions.
“The RightToRead program gives us the opportunity to focus on enabling English literacy and proficiency among the under-served populations of India. Intel and AWS provided the scalability and adaptability through the cloud to extend the program to the remotest parts of India and make a difference for both students and teachers,” says Sanjay Gupta, Global CEO, EnglishHelper.
EH leveraged the AWS Elastic Compute Cloud (EC2) M5 instances powered by 2nd Generation Intel Xeon Scalable processors with Intel Deep Learning Boost (Intel DL Boost) to seamlessly deploy the AI-based RightToRead program across schools in urban and rural parts of India. Intel processors and architecture employed to support the AWS application servers and databases helped improve compute times and ensured optimisation of the applications. The Relational Database Service (RDS) that AWS provides, in addition to Intel Performance Libraries, helps with setup, operation, and scaling of the database used for the application. The Intel AVX-512 extension along with AWS content delivery networks (CDNs) greatly enhance the performance of EH’s machine learning algorithms and improve application availability and performance.
Seamless deployment using Intel and AWS
One of the major challenges that EH faced was deploying the software to multiple schools simultaneously. Working with Intel and AWS, EH could migrate the solutions to an AWS cloud from their on-premises data centre. Intel architecture and AWS cloud provided scalable computing capacity, keeping costs low and enhancing the learning experience. With automated deployment, QA solutions, and stateless servers—made possible using Amazon EC2 M5 instances based on Intel Xeon Scalable processors—EH was able to bring down deployment times from four hours to 30 minutes and planned downtimes for every release from four hours to a few minutes. Moving to the cloud also ensured that the installation and registration of the software could be done in under 30 minutes.
Another challenge that EH faced was providing effective operational support for all the schools that were using the programme, especially in remote locations. The need to physically visit each school and limited information provided about the functioning of the software meant increased operational costs and a larger team size. Working with Intel powered AWS instances and cloud helped EH manage the infrastructure with a team of just two people while significantly lowering their capital expenses. With most routine tasks like provisioning of servers, scaling in response to demand, and healing being managed by Intel and AWS, EH could focus on improving their infrastructure and solving real customer issues.
Utilising Intel architecture and AWS technology, EH can now achieve uninterrupted scalability and implement their solutions in more than 100,000 schools across India. This solution will effectively help EH make a lasting impact on the lives of over 20M students and 500,000 teachers, enabling them with improved learning abilities and promoting continual growth and progress.