Situation: I need to make a website for my Final Year Project. It's the website for brain analysis. The website features include chat, blogs, posts, users, payment methods. One of the main features includes the use of AI, which I know only in Python.
Decisions and Confusions: I decided to make two backends and one front-end. One backend will be using Django with GraphQL/RestAPI that will be running my AI models. The other backend is for the website. It will add users, chat, post, etc. I'm thinking of using TypeScript, Prisma, ExpressJS, GraphQL, MongoDB/PostgreSQL.
Please guide me to the latest and stable tech stack I can use. Because one of the requirements of our Final Year Project is to use the latest tech stacks. 1st Backend advice? (This will be used to run AI models) 2nd Backend advice? Frontend to 2nd Backend advice?
Thank you.
Hey there 👋,
Daniel from the Prisma team here.
I think your choice of a stack would work well for your final year project.
Some recommendations: - Use PostgreSQL if you need a stable stack. Prisma support for MongoDB is currently in Preview and therefore isn't stable. Moreover, PostgreSQL being a relational database enforces a schema more strictly than MongoDB which is useful given that your data model involves multiple relations. - If your Django backend exposes a REST API, you can also expose it over the GraphQL API by proxying requests from the GraphQL API to the REST API. That way, you have a unified API for all operations. This is typically known as wrapping. - Regarding the GraphQL part, I would consider looking at Nexus and nexus-prisma.
For inspiration, check out the Prisma Examples ​repository which contains many ready-to-run examples.
Here's another fully-fledged example using Prisma, Fastify, GraphQL, and PostgreSQL: https://github.com/2color/fastify-graphql-nexus-prisma
Hi. Maybe you can try use FastAPI instead Django https://fastapi.tiangolo.com It could be faster. The FastAPI documentation is so useful and elegant.
Also you can try split a little more the backend and use an "microservice" architecture. Using Kubernetes to deploy your services.
