Inspiration
ChatGPT often gives us wrong results because we don't use proper prompts when asking ChatGPT anything. We have Gathered some high quality prompts from certain categories which will produce high quality results.
What it does
Chatur is a Web App which gives you human like text based on certain predefined prompts which are integrated into the application. Before deciding to include them, we investigated various tools, such as AIPRM, and tested a number of high-quality prompts. End users find it simpler to produce high-quality results without having to worry about prompts themselves.
How we built it
We created the entire application in NextJS, and we created the API using the NextJS API directory, which eliminated the need to create a separate backend server. To fully utilise Vercel's infrastructure, we used NextJS' Edge Runtime. The edge functions made our API calls extremely quick.
Challenges we ran into
It was new because we had never used OpenAI before. We also encountered a GPT 3.5 Turbo token size error because we were consistently exceeding the maximum size of 4,096 tokens. To fix it, we decreased the token size in our configuration.
Accomplishments that we're proud of
We had never participated in a Hackathon before so it was an accomplishment for us to build the entire project in the span of 48 hours.
What we learned
We learnt about OpenAI, Edge Runtime and NextJS API Routes.
What's next for Chatur
We will add more high quality prompts in the future and keep on improving it.
Built With
- chatgpt
- headlessui
- nextjs
- openapi
- react
- tailwindcss
- typescript
- vercel
Log in or sign up for Devpost to join the conversation.