Hackathon for Social Impact: Transforming Education with AI
Pollyana Azevedo
By: Pollyana Azevedo.
Introduction
"How can we create space for and foster fast, disruptive innovations at Fundação 1Bi?"
Well, we are a nonprofit organization with the great dream of bringing technology to make the daily work of teachers in Brazil’s public K–12 education system easier. However, how can we combine our know-how, our desire, and the engagement of the AprendiZAP team to innovate and thereby expand access to emerging technologies completely free of charge?
That is how we decided to bring the entire team to a halt and hold the first Hackathon at 1Bi, something like a marathon that brings specialists together to create innovative solutions in a short period of time. This alone brings a certain boldness that is not so easy to implement amid the constant discovery and delivery routine of our existing products. But at 1Bi, we like to be bold.
The Hackathon brought together our innovation, content, design, data, and engineering teams to explore and revolutionize how Artificial Intelligence is used in educational settings. We started with a big idea: how can we use AI assertively to support, rather than replace, the hard work of educators?
The Hackathon schedule was structured to last 10 business days, and we divided the team into two groups. During the first 5 days, we blocked everyone’s calendars for a full immersion in the challenge of understanding the various uses and trends in AI. We studied Artificial Intelligence, education, and how to connect these two major areas based on our mission, product vision, and values. We held brainstorming, decision-making sessions, and, at the end of the first week, each team brought forward a proposed solution prototyped by the designers. The following week, the technology and engineering team focused on developing, testing, and implementing each team’s solutions. Below, I will describe the hackathon, the methodologies applied, and the results achieved. Will you join me?
Methodology of the Hackathon
To make the Hackathon possible, an important part was structuring a methodology. In this regard, two points were important: making good use of the team’s available time while introducing a format in which the knowledge acquired during the first days could be put into practice.
Thus, we combined the Design Sprint and Challenge Based Learning. By integrating these approaches, the agile structure of the Hackathon would be maintained through the Design Sprint, while emphasizing in-depth investigation for solutions to real social problems, characteristic of the framework CBL.Below, I explain the approaches and how we integrated them:
Design Sprint
Design Sprint 1.0 by Jake Knapp
With the methodology developed by Jake Knapp and made renowned by Google Ventures, the Design Sprint could help us map, ideate, and prototype solutions during the first half of the hackathon. We adapted the methodology to our established routine and culture and used it as a guide for our actions. Thus, we divided it into five stages: Understand; Sketch; Decide; Prototype; Test.
Challenge Based Learning (CBL)
Challenge Based Learning by Dublin City University (DCU)
With the CBL methodology, whose name translates into Portuguese as Aprendizagem Baseada em Desafios, the intention was to propose collaborative solutions to relevant social challenges. Participants are encouraged to reflect and formulate essential questions and key issues to frame the challenge, and throughout the process acquire knowledge that will help shape proposed solutions. CBL is divided into three main phases: Engage; Investigate; Act.
Structure of the Hackathon
Then, moving on to structuring the hackathon, I looked for similarities between the two approaches, which share common principles in the design process. Combining the Design Sprint and CBL enabled a holistic approach to the hackathon in which we agilely considered the social impact of the solutions. In the schedule, we divided it into four main phases:
Understand and Define (Days 1 and 2):
Using the initial idea from the CBL methodology and with the aim of bringing participants’ knowledge to the same level, the first day was set aside for researching the use of AI in education. This topic served as the starting point for the hackathon, that is, the Big Idea proposed by CBL.
Thus, through the Design Sprint, the teams began by identifying real problems in education and new technologies to establish essential connections that would guide the hackathon. This initial stage ensured that all solutions were relevant and centered on users’ actual needs.
Finally, everyone was able to bring references, studies, and insights on the topic so that each squad could define the solution’s objectives, metrics, and risks.
Generate and Refine Ideas (Days 2 and 3):
To generate ideas, we applied brainwriting techniques such as the 365 Method and rapid-sketching techniques such as Crazy 8s. The first technique enabled full participation from everyone on the team, while the second allowed even those who did not draw to sketch and visualize complementary ideas.
The teams discussed and voted on the best ideas, creating wireframes and planning prototypes.
Prototyping and Testing (Days 4 and 5):
At this stage, participants built rapid prototypes and conducted research with real users. The feedback obtained was essential for adjusting and refining the solutions.
Development and Iteration (Days 6 to 10):
From this stage onward, the focus was on the engineering team. Continuing with the iterative approach of the Design Sprint, the teams developed and refined their solutions.
The CBL approach continued to guide the investigation and data collection, ensuring that the solutions were relevant to our users and could be put into practice and launched for use at scale.
Metrics and Results
Project Highlights
The projects developed addressed various aspects of education with AI, including pressing topics such as personalized learning, inclusive education, automation of administrative tasks, and gamification of teaching. Each team presented its project in a pitch, highlighting its main features and expected impact. One of our solutions was to use Artificial Intelligence to give "life" to Mari, AprendiZAP’s virtual assistant. Thus, through good practices and the assertive use of this technology, we can support our teachers with high-quality pedagogical assistance. This solution can be used both on the AprendiZAP Professores platform itself and via WhatsApp at (11) 3230.2918 to create lesson plans, answer teachers’ questions, generate personalized content, and for many other educational purposes.
As a second solution, we considered experimenting with the Records feature to record information about classes, making it easier to produce reports and insights about classes and students.
Conclusion
The first edition of the Hackathon at Fundação 1Bi was a transformative experience that demonstrated the power of interdisciplinary collaboration and agile methodologies in creating innovative solutions for education. The methodology provided a structured yet adaptable process, ensuring that the solutions developed were practical, relevant, and impactful.
As a result, the format was highly successful: we achieved 100% satisfaction with the proposed solutions and received positive feedback through an anonymous survey of team members.
The icing on the cake was also reflecting this structure in our day-to-day work. We chose to maintain squads A and B so that we could continue fostering this participatory, collaborative, and interdisciplinary approach. Thus, each team can take ownership of different products and projects, collaborating both with one another and across squads.
Next Steps
The most promising projects will receive support to continue their development and implementation. In addition, we are continuing to structure the next hackathons, analyzing and applying lessons learned to foster ongoing innovation in education through the use of emerging technologies such as artificial intelligence.
Pollyana Azevedo is design lead at Fundação 1Bi.
