A Nextios, LWSA company, developed a Digital Teacher based on AWS AI for Education – Technology and Innovation Ecosystem, business area of Positivo Tecnologia. The solution was presented during Bett Educar 2024, innovation and technology event for education, carried out in April of this year. The Digital Professor uses an architecture based on AWS services like Amazon Bedrock, AWS Lambda and Amazon Transcribe.
The Digital Professor offers quick and effective answers, enhancing the user experience. In addition to summarizing classes and preparing exams, the solution has the ability to generate content automatically. The goal is to make the Digital Teacher widely available to schools, improving educational management and the learning experience
Experimenting with AI as an educational tool
In addition to your product portfolio that ranges from computers, smartphones the payment machines, Positivo Tecnologia has a business area focused on developing solutions for education, like the Educational Hub. Among the solutions available in the Hub, the Aprimora platform gained a generative artificial intelligence assistant last year – to Maria – to help students understand specific issues and answer them
According to Christiano Grillo Justus, head of Engineering and Architecture of Education, from there, in 2023, conversations with Nextios have begun, AWS partner, for the use of generative AI, arriving at the Digital Professor project, integrated solution to the Educational Hub.
The first concept was built in theAmazon Bedrockwith content from the BNCC (National Common Curricular Base) and various teaching materials."In this stage, she learned from these materials and, by the user's age, "he presumed what he should know by offering information in the tone of voice of the schools associated with the system", explain Justus
According to Carlos Alberto Marangon, technical leader and AWS ambassador at Nextios, the conversations with AWS led to the development of a new architecture model, until then not used. "The idea arose to use the Amazon Bedrock Agents only in the final stage", to bring the answers from the knowledge base, instead of developing the entire solution within the service, explain.
The teams created a service-based architecture such asAWS Step Functions, AWS Lambda, Amazon OpenSearch Service, Amazon EventBridge, Amazon Transcribe, Amazon CloudFront, Application Load Balancer, AWS Certificate Manager e Amazon Simple Storage Service (Amazon S3). Already on Amazon Bedrock, there is the use of Anthropic Claude-3, Knowledge BasesandAI Agents.
For Gabriel Leite da Silva, AWS Partner Solutions Architect, the implementation strategy – isolated in specific tasks – allowed the balance between generic information, that should not be used; and very specific, that would make the solution similar to chatbots.
The specialist emphasizes the use of a technique called prompt chain, that mixes low code solutions with the code envelope that allows the final response to the user to be made through stream, what allows combining business flow management with very quick responses from Amazon Bedrock.
"This combination of services is a paradigm shift". We opted for a serverless architecture and AWS tools that provide this low code part, allowing not only to deliver quickly, but giving Positivo Tecnologia the opportunity to experiment with different scenarios, affirms.
"She is not only focused on a knowledge base", being able to call a systemic API, understand your output and perform a systemic processing without the need for code to build these calls, affirms, highlighting the possibility of integration with transactional and analytical workloads, with low code usage.
Meeting demands and carrying out tasks
Marangon, from Nextios, reinforces that the architecture, how it was designed, allows to meet the demand and perform tasks of different user levels, with access to specific content for teachers, students and managers.
The use of Amazon Transcribe, for example, allows recorded classes to be saved and transformed into summaries by Amazon Bedrock that can later be accessed by students. "The recorded class becomes a searchable content for the student", explain, remembering that the solution, with these components, made things that were not visible viable and brought a performance that was not possible before.
Justus claims that the first proof of concept was developed in two months. Confirmed the operation, a second version was prepared in 15 days, allowing the Digital Teacher to be presented at Bett Educar 2024. In this phase, the integration with the solutions developed by edtechs and offered in the Educational Hub was included. In this way, when consulted, the solution indicates that materials can be used within the existing solutions in the Hub
"To introduce the Digital Professor", we created a video of a class, showing that he can understand a video, allow students to ask questions and teachers to create a more efficient study plan based on that class, specific for certain students. Furthermore, the student could ask questions about the material and clarify doubts about the explanation, "also based on the knowledge accumulated by AI", explain
Justus compares one of the functions of the Digital Teacher to that of a teaching assistant. In the future, the director of a school who wants to know how his classes are doing, you will be able to view this information and receive management suggestions. It is an action assistant, feedbacks, data and material. "That's where we're going", say
The next step, according to Justus, is to present the solution to the business team. The objective of this process is to define the path for delivering the greatest value in the shortest time, bringing the solution to schools that can actually test them.
André Amorim, general director of Nextios, reinforces that "This project represents the first step of many that we are taking in building solutions with Generative Artificial Intelligence. The Digital Professor is not just a project, it is a reflection of our commitment to remain at the forefront of technology and to create solutions using the most modern resources for our clients, affirms.
