Summary
AI Tutor aimed to expand Brainly's potential to answer any question on any subject that a student might have, connecting this new solution to the company's premium plan that counts with specialized tutors.
The idea was that students could start the flow by asking AI first. During a session, they would be able to get tailored answers at any time and if they still need further explanation, they would be able to connect with a specialized Brainly tutor. Those professionals would leverage the answers given by AI to elaborate more on top of it.
By doing that we believed we could expand Brainly's potential for answering questions related to more subjects. And also Brainly would be able to provide answers to a wider variety of questions (levels) within each subject, with greater precision and speed, thus reducing operational costs to the company, among other benefits.
The project progressed through various stages, ultimately culminating in a successful first release to a subset of users (20% on web) by the end of Q2 in 2023.
By doing that we believed we could expand Brainly's potential for answering questions related to more subjects. And also Brainly would be able to provide answers to a wider variety of questions (levels) within each subject, with greater precision and speed, thus reducing operational costs to the company, among other benefits.
The project progressed through various stages, ultimately culminating in a successful first release to a subset of users (20% on web) by the end of Q2 in 2023.
.
Company • Brainly - Ed-tech
Role • Senior Interaction Product Designer
Timeline • 2023 - Q2
Metrics and Impact
Results gathered in this project
Context and problem to be solved
1. Company overview
Brainly is a leading online learning platform, connecting students with peer support, expert help, and advanced learning tools. The platform’s mission is to democratize education by making it accessible to students around the world.
2. What we were trying to solve
Brainly wanted to expand its ability to answer a broader range of questions across different subjects with greater precision and speed.
Traditional support methods were resource-intensive and costly. To address this, we introduced an AI-powered solution that could handle a wide variety of questions, reduce operational costs, and enhance user engagement.
Traditional support methods were resource-intensive and costly. To address this, we introduced an AI-powered solution that could handle a wide variety of questions, reduce operational costs, and enhance user engagement.
3. Let’s take a closer look
The existing 'Ask an Expert' feature was effective but had limitations in both scope and speed, often leading to delays and incomplete responses for users. By integrating an AI-driven solution, Brainly could deliver faster, more accurate answers while reducing the dependency on human effort alone.
Objectives - What we set out to improve
1. Innovative learning solution
2. Enhanced user engagement
3. Seamless web integration
4. Positive user feedback
2. Enhanced user engagement
3. Seamless web integration
4. Positive user feedback
1. Innovative learning solution: Introduce a new AI Tutor feature to address a wider range of student inquiries with speed and precision.
2. Enhanced user engagement: Boost engagement by offering a new, more efficient way for users to get answers.
3. Seamless web integration: Ensure that the AI Tutor is easily accessible and smoothly integrated within the existing Brainly platform.
4. Positive user feedback: Achieve high user satisfaction through effective and helpful AI-driven responses.
2. Enhanced user engagement: Boost engagement by offering a new, more efficient way for users to get answers.
3. Seamless web integration: Ensure that the AI Tutor is easily accessible and smoothly integrated within the existing Brainly platform.
4. Positive user feedback: Achieve high user satisfaction through effective and helpful AI-driven responses.
Role - Senior User Interaction Designer
Responsibilities
Tools and Techniques
1. Desk research
2. Prototyping
3. Design QA
4. A/B testing
2. Prototyping
3. Design QA
4. A/B testing
1. Desk research: Conducted research to gather insights on user needs, market trends, and AI-driven educational tools.
2. Prototyping: Developed interactive prototypes to visualize and iterate on the AI Tutor feature. They were used in stakeholder/ C-level presentations and also help on guiding developers during implementation.
3. Design QA: Conducted thorough quality assurance checks to ensure the final product adhered to design specifications and delivered a consistent, high-quality user experience.
4. A/B testing: leveraged A/B tests to compare different versions of the AI Tutor feature vs the current solution.
2. Prototyping: Developed interactive prototypes to visualize and iterate on the AI Tutor feature. They were used in stakeholder/ C-level presentations and also help on guiding developers during implementation.
3. Design QA: Conducted thorough quality assurance checks to ensure the final product adhered to design specifications and delivered a consistent, high-quality user experience.
4. A/B testing: leveraged A/B tests to compare different versions of the AI Tutor feature vs the current solution.
Desk Research - Uncovering industry insights
1. How ChatGPT will impact the ed-Tech industry?
According to an article published by APN NEWS (Feb 2023) one of the critical areas where ChatGPT could have a significant impact is tutoring and personalized learning.
With the ability to comprehend and respond to questions in real time, ChatGPT could be used to provide students with real-time, on-demand tutoring services. This could include answering questions about specific subjects, providing additional explanations and clarifications, and generating personalized learning plans based on a student’s strengths, weaknesses, and learning style.
Additionally, ChatGPT’s ability to provide relevant and accurate information could help to address the challenges of access to educational resources in under-resourced communities.
By providing students with access to information and educational resources that they may not otherwise have access to, ChatGPT has the potential to help level the playing field and provide students with equal opportunities to succeed.
With the ability to comprehend and respond to questions in real time, ChatGPT could be used to provide students with real-time, on-demand tutoring services. This could include answering questions about specific subjects, providing additional explanations and clarifications, and generating personalized learning plans based on a student’s strengths, weaknesses, and learning style.
Additionally, ChatGPT’s ability to provide relevant and accurate information could help to address the challenges of access to educational resources in under-resourced communities.
By providing students with access to information and educational resources that they may not otherwise have access to, ChatGPT has the potential to help level the playing field and provide students with equal opportunities to succeed.
2. Global AI in the education market revenue by region
The most impacted region (North America, as the graph below shows) includes the USA, where Brainly Tutoring focused its operations.
Process & Collaboration
1. What could be improved
• Personalization: Further enhance the AI's ability to provide tailored responses based on individual learning styles.
• Scalability: Plans to scale the AI Tutor to more users without compromising performance.
• Scalability: Plans to scale the AI Tutor to more users without compromising performance.
2. What worked well this time
• User engagement: The AI Tutor feature successfully engaged users, with significantly higher interaction rates compared to previous solutions.
• Design-development collaboration: Strong collaboration ensured a smooth and efficient integration of the AI Tutor into the platform.
• Design-development collaboration: Strong collaboration ensured a smooth and efficient integration of the AI Tutor into the platform.
Designs
Ready for development
Learnings
1. What did I learn or love in this project?
.
• Exploring AI in education: It was exciting to work on an innovative solution that could significantly enhance the educational experience for students.
• User-centric design: Quickly validating our designs through A/B testing was crucial in ensuring the feature met real user needs.
• User-centric design: Quickly validating our designs through A/B testing was crucial in ensuring the feature met real user needs.
2. What could have been done differently?
• Better alignment between key-stakeholders: At this point, more than one AI-oriented initiatives were happening in the company in parallel and it they were overlapping without the crucial alignment between teams.
• Anticipating scalability challenges: While the feature was successful in the initial rollout, better planning for scaling could have helped mitigate some of the post-launch challenges.
• Refining the feedback mechanism: Improving how we collected and acted on user feedback could have provided more actionable insights early on.
• Anticipating scalability challenges: While the feature was successful in the initial rollout, better planning for scaling could have helped mitigate some of the post-launch challenges.
• Refining the feedback mechanism: Improving how we collected and acted on user feedback could have provided more actionable insights early on.