feat(agent): Enhancing the overall application layer

[EncHawk]

The need for random decision making and human in the loop feedback has only proven to be beneficial.
With the possible inclusion of diffusion in this pipeline, its only better to have nodes that provide non-linearity instead of a simple langchain workflow. This issue will end up tracking all other issues that are related to this.

Besides, this introduces tools for web search and using a multi-tenant vector database for storing context is better.

• Replacing Langchain code, setting up langgraph with the dependencies: feat(agent) : replacing langchain with langgraph #1
• Enhancing RAG pipeline with web-search and a full-fledged vector database : feat(RAG): integrating web-search and vector databse #2

more features will be tracked here

[EncHawk]

WE COOKED UNBELIEVABLE HOW MY BRAIN BABY HAS COME THIS FAR. im proud of muself. gotta do the frontend now.