Inspiration
As a A level Student, I always have little time to mark my papers after I done countless of them. That is when I decided to create my own automated solution. This way I can focus more on real time feedback and sharpen my exam skills.
What it does
PinchAI takes away your workload by automatically marking your papers, tracking your progress and performance, and being able to become your real time tutor, accelerating your progress.
How I built it
Using TiDB features and capabilities from vector search to normal queries, I was able to integrate a suitable pipeline for an autonomous workflow using OpenAI LLMs.
Challenges we ran into
Planning is most definitely the hardest part of this project. I had to scrap multiple approach as some of my stacks are just not compatible or very expensive in the long run
Accomplishments that we're proud of
Being able to do what it is told with acceptable accuracy, allowing students to take off their time off marking their own papers.
What we learned
There are lot of challenges when working with databases such schemas, limits and the complexity of vector search and vector stores.
What's next for PinchAI
The next step for PinchAI is to truly develop further enhancing its features to be faster, easier and scalable.
-- tidb associated account
Built With
- nextjs
- openai
- tidb
Log in or sign up for Devpost to join the conversation.