We were inspired to create Findingo after seeing and experiencing the frustration of using archaic textbooks with no accessibility on the web. Especially in our computer science classes, we would spend lots of time transcribing code from textbooks, and these difficulties inspired us to create Findingo.

What it does

Findingo bridges the gap between the traditional print medium and modern digital resources using NLP, OCR, and additional technologies to provide a summary of the textbook page, definitions for key terms, links to assorted online resources, and automatic code analysis. These methods allow for transferring all the knowledge on pages of textbooks to the modern age.

How we built it

Our application uses a novel 2 layer, 4 tier architecture. The user interacts with an intuitive and user-friendly web-based interface that uses Material UI and is powered by React. The data is sourced from a Rust processing engine, which is compiled to WebAssembly and is thus able to efficiently run entirely in-browser, for the optimal user experience.

Challenges we ran into

During the creation of Findingo we faced many challenges. One challenge was that our machine learning based summarizer was too inaccurate, and provided a very simplistic summarization of the text. To rectify this, we utilized the bleeding-edge GPT-3 summarizer which suited our needs.

Accomplishments that we're proud of

We’re proud of being able to create and deploy multiple fully functioning AI models into a production ready environment in less than 48 hours. We’re also proud of the intuitive user interface and experience we were able to build with Findingo.

What we learned

We learned a lot about optical character recognition and machine learning. Using GPT - 3 gave us a lot of insight into the ability of AI to perform tasks accurately and efficiently.

What's next for Findingo

We plan on adding integrations with online code IDEs such as, and creating a mobile version of Findingo.

Share this project: