Recommending the perfect book

Description

As the lead front-end developer on the Verhalenvinder project, we were tasked with creating a coded prototype of an AI-powered recommendation app. The project's aim was to address the issue of low reading skills among youth in lower education schools, building upon previous qualitative research done with the target audience. Over a four-week period, I used a Scrum-ban approach to manage tasks, ensuring the project was delivered on time and within budget. I also implemented a measurement plan in Google Analytics (GA4) for monitoring during the pilot phase.

Business and user outcome

The Verhalenvinder pilot phase was very successful for both business and user. Through the use of Verhalenvinder, students were able to discover books that interested them and improve their reading skills. For the business, this led to an understanding of the process that students go through when selecting a book. The Verhalenvinder pilot reduced innovation risk for the business and provided valuable insights into their target audience. Clear next steps were identified based on these insights..

Biggest Challenge

The absence of a book database and the stakeholders' limited technical knowledge posed significant challenges. We solved this with clever use of a custom webscraper and enriched this data using OpenAI and Langchain. I kept the stakeholders updated on all significant steps in a non-technical language to manage expectations effectively.

What I've Learned

This project allowed me to gain experience with Langchain and OpenAI for data enrichment and recommendation calculations. I also improved my knowledge of GA4 and Nuxt 3. Importantly, it enhanced my ability to translate business needs into technical implementation when working with a non-technical team.

Hard Skills

• Nuxt 3 / Vue 3
• TailwindCSS
• Typescript
• Vercel
• GA4
• Hotjar

Soft Skills

• Communication: Effectively conveying technological advances to a non-technical audience.
• Expectation Management: Managing project scope and stakeholder expectations, particularly given the project's nature as a coded prototype rather than a full product or MVP.