Our company has a mobile app that recommends relevant digital content filtered from social media, to help parents nurture their children's curiosity.

The website welcomes parents, educators, therapists and trusted voices from the community to curate and share different content templates to help kids achieve specific learning goals.

We would like a group of students to improve the AI models driving our current recommendation engine. These recommendations must be able to draw from even more social media platforms than we currently have, and flag both the relevance of content to the child's curiosity, and any relevant information about the views expressed in the videos.

Finally the model should also be able to guide parents through the next steps they can follow to ground digital learning in actual life experience/tasks.

This will involve several different steps for the students, including:

• Familiarizing yourself with our website and product to understand how they work.
• Researching state-of-the-art machine learning and content/feed curation technologies.
• Developing models, and improving our current evaluation dataset used to test them using (langchain/python)
• Testing recommendation engine prototype models with users.
• Iterating and improving tested prototype models.

By the end of the project, students should demonstrate:

• Understanding of our website and products
• Understanding of machine learning and content/feed curation technologies
• Production of recommendations using prototype models

Bonus steps would include:

• Testing prototype models with users and iterating designs to improve them

Final deliverables should include

• Code developed for the recommendation model.
• Recommendations supporting the model’s validity, which may be assessed subjectively.
• A written report describing the technologies that were used, the challenges that were faced, the model’s final outcomes, and recommendations for the project’s next steps.