Blob: AI-Teaching Assistant

An AI powered Teaching Assistant that allows educators to create an knowledge-based repository of the course content, which enables a gamified personalized student-centric learning experience.

Comment

Inspiration

With emergence of new technologies we see a radical shift in industries, we wanted to explore the capabilities of AI to accelerate the learning experience.

What it does

Blob allows educators to create an knowledge-based repository of the course content, which enables a gamified personalized student-centric learning experience.

How we built it

Built with NextJS, which serves the API endpoints and the frontend client, internally pushes uploaded documents to Pinecone as embeddings and invokes OpenAI's GPT-4 API with the available context and queries to generate responses for the user. (love and dedication were the secret tools)

Challenges we ran into

  • Pruning available knowledge to build the right context for queries.
  • To narrow down the project scope to focus on one problem at a time.
  • Design a business model that can be scaled.

Accomplishments that we're proud of

We were able to implement a scalable AI assistant that can compete with proxies like DanteAI, and implement more features than we planned. We were also able to develop key insights on customer acquisition, adoption and retention strategies.

What we learned

Technical: A lot about Vector DBs, pain points in the current education industry, and business ideas in education domain. Business: What thoughts and concerns should be kept in mind while making a product.

What's next for Blob: AI-Teaching Assistant

  • Implementing integrations for LMS and LTI's.
  • Building a student view to facilitate AI-enabled learning based on the Instructor's knowledge base.
  • Create a shared knowledge based repository.
  • Expand learning for purposes not limited to academia

Built With

Share this project:

Updates