Things to learn -

1. Interacting with LLM via APIs
2. HTTP Responses vs streaming responses
3. Websockets to stream back response to the user
4. MultiModal Inputs using URLs
5. MultiModal inputs by uploading a file
6. Add multiple API providers (openai, grok, claude etc)

https://ai.google.dev/gemini-api/docs

https://aistudio.google.com/apikey