UX Design
Instant Answers
Lead Product Designer & Researcher
I designed the user experience for a new AI-aided service, collaborating closely with the Business and Product Owners, Machine Learning team, and both backend and front-end developers. I successfully led the team to deliver the solution and conducted testing to validate its value.
Background
On Brainly, a platform where students ask and answer questions, around 50% of queries were duplicates. This led to numerous new answers to old questions, without improving the overall knowledge base. Consequently, original questions received slower responses, significantly impacting user satisfaction, as young users typically expect instant answers.
~50%
questions asked on Brainly are duplicates
Solution
Working closely with the team, we developed a service that leverages machine learning to identify the best answers from the knowledge base, considering factors like view count, votes, and “hearts.” As users typed their questions, the service checked in real-time to see if similar questions had already been answered on Brainly, providing them with high-quality answers instantly. This approach allowed resources to focus on addressing unique questions and enhancing the overall knowledge base.
Test results
The concept was proven during A/B/C/D tests on US and Polish markets.
Few years after inroduction the solution is still live and continues to help 350 million students on Brainly.
90%
of less duplicates on markets with Instant Answers service introduced
25%
faster answers to unique questions