langchain-pdfbot/README.md

A LangChain-based tool to answer questions strictly from PDF content using Groq LLM and local embeddings.

# 📄 LangChain PDFChat

**LangChain PDFChat** is a lightweight Node.js library that helps you build a local question–answering chatbot over any PDF.  
It uses local embeddings (so you don’t pay per token) and powerful Groq LLM to generate answers **strictly based on the PDF content**.

✅ **Only answers questions if info is found in the PDF**  
✅ **Fast**: Uses in-memory vector search  
✅ **Easy to use**: Just point to a local PDF, ask questions

---

## ✏️ How it works

1. You load a PDF file from your local system.
2. The tool splits & embeds the PDF content in memory.
3. A retriever fetches only the relevant chunks for each user question.
4. The question and chunks go to the Groq LLM to generate an answer.
5. If the answer isn't in the PDF, it says: `"Not relevant question"`.

---

## 🚀 Quick Start

### Step 1: Install

```bash
npm install langchain-pdfbot
```

## Step 2: Get your free Groq API key
- Sign up at: https://console.groq.com
- Go to **Settings → API Keys → Create API Key**
- Copy your new API key (starts with `gsk_...`)

## Step 3: Add your API key to .env
GROQ_API_KEY=your_groq_api_key_here

## Step 4: Use in your code
```bash
// index.js

import {createLlm} from "./model/llm.model.js";
import {createPdfQaTool} from "./tools/pdfQa.js";

const llm = createLlm({ apiKey: <groq-api-key>, model: "llama3-70b-8192" });

const askQuestion = await createPdfQaTool("./pdfs/sample.pdf", llm);

const answer = await askQuestion("what is array? ");
console.log("Answer:", answer);

```