Movie-Trivia LLM

I’m a huge fan of Star Trek. Since I’m hooked, I dive deep into theories, taglines, interpretations, and how the stories evolve.

Ever thought about building an app that brings all those fan theories together in one place? Powered by AI, and written in just 1K lines of code? It’s easily attainable.

“Hello World” apps are a thing of the past. Why not blend your interests with the tech stack you want to master and the user experience you want to create?

In essence, you can gather your favourite stories, feed them into an AI engine, and start asking questions that spark deeper insights.

Quick Presentation

What parts do I need?

Hmm… what have I just done here?

Let’s dissect. I chose:

ReactJS because, for me, it is quicker to build the frontend. You can choose the one you like;
Python has a great and active community. I need LlamaIndex to build my knowledge base;
OpenAI to generate text embeddings;
Milvus vector database is easy to integrate. Out of all the features it provides, it lets you configure multiple vectors. I don’t need this in this project but it’s good to notice Milvus supports it;

What do I have to develop?

Hmm… good question

Let’s list them down.

You will need a dedicated folder where you can put all your sources of Star Trek;
A sequence of steps to parse and feed documents into Milvus;
An API (probably, WebSockets) to return summarization when a user asks;

Here is the quick overview of what it would look like:

That’s it?

For now, yes. If you would like to know how I built this project, do drop an email from Contact. I will plan to release a learning material around this.