Didn’t know I had 7 hours stored in here, but hey, free time!
This is a port of an app I made a few years ago to next.js so it is deployable. This is a simple (or complex, depending on how you look at it) citation machine. It is powered by 5 different services in the backend: chegg, mybib, readability, hackclub ai, and my own parser. They all work together to ensure credibility and correctness in citations based on a URL. Wait? AI? Doesn’t that hallucinate????!?!?! No. Not the way I did it. The model is constrained to a specific schema and will only change things if most of the previous sources disagree on eachother (old data from a random API is usually less reliable than data scraped off the page by an LLM). Backend does all the fetching, client privacy through a proxy.
0
Log in to leave a comment