So, I'm learning how to build LLM-powered chat apps that are more production-ready, from scratch.

I've created an open-source repo that contains my ongoing explorations.

The stack:
• Django backend
• React frontend
• @LangChainAI agents
• Websocket protocol

GitHub below. In my free time, my main focus will be to evolve this codebase.

Functionalities I want to add:
• Streaming LLM output
• Storage (e.g. vector db)
• More agent tools
• Short-term memory
• BabyAGI
• Plan-and-Execute agent
• More sophisticated UI
• Error handling

What else?

Alright, I finally understand @LangChainAI agents and tools.

I can now create a custom:
1. Tool that "reads" annual reports
2. Agent that answers queries via the tool

For my example, I am using $META's 2022 annual report.

Code is below. Happy learning 🙂 (1/3) About the custom Tool

Under the hood, the Annual Report tool is forwarding queries to a @LangChainAI chain.

This chain answers queries using:
• @trychroma's Chroma DB
• @OpenAI's GPT-3.5 Turbo model

I still don't have API access to GPT-4. Sigh.

I've seen multiple questions about how to build a Chatbot that:

• Retrieves data from PDFs and
• Has conversational memory

Turns out, it's really simple to do with @LangChainAI.

So, I wrote a quick tutorial with a real-world example for you all.

Code below. 1/4

Prompt: "What was Airbnb's net income in 2022?"

Response: "$1.9 billion"