File size: 2,176 Bytes
92b7244
 
 
81693a8
 
92b7244
81693a8
92b7244
 
 
 
ac013c2
c494433
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
---
title: Chat With Documents
emoji: πŸš€
colorFrom: purple
colorTo: purple
sdk: streamlit
sdk_version: 1.13.0
app_file: app.py
pinned: false
---



---

# Chat With Documents πŸ€–πŸ“„

Welcome to the **Chat with Documents** app! πŸš€ This Streamlit app allows you to upload PDF and PPT files, extract their content, store the extracted text in a vector store, and interact with it using natural language queries! πŸ€–πŸ’¬

Built with **LangChain**, **OpenAI**, **Streamlit**, and **Astra DB**, this project leverages the power of LLMs (Large Language Models) to allow users to chat with their documents like never before. 🧠

---

### πŸš€ **Features**

- **PDF & PPT Extraction**: Upload PDF and PowerPoint files to extract text! πŸ“„βž‘οΈπŸ“
- **Vector Store**: Automatically stores extracted text in a **Cassandra** vector store. πŸ”πŸ“š
- **Ask Anything**: Ask questions about the document and get answers powered by **OpenAI**! πŸ€–β“

---

### πŸ› οΈ **Tech Stack**
- **Streamlit**: Frontend framework to interact with the app.
- **LangChain**: For seamless document processing and querying.
- **OpenAI**: For LLM integration to provide intelligent responses.
- **Astra DB**: Database for storing and managing vectorized text data.
- **Python Libraries**: PyPDF2, python-pptx, cassio, and more.

---



### πŸ’‘ **How It Works**

- Upload a **PDF** or **PPT** file using the file uploader. πŸ“€
- The app will extract text from the file using **PyPDF2** (for PDFs) or **python-pptx** (for PPTs). πŸ“„βž‘οΈπŸ“
- The extracted text is split into manageable chunks using **LangChain's CharacterTextSplitter**. βœ‚οΈ
- The chunks are then added to **Cassandra** as vectorized data using **OpenAI embeddings**. πŸ”„
- Ask any query about the content of your document, and the app will respond using the power of **OpenAI**! πŸ€–πŸ’¬

---

### 🎯 **Why Use This?**

- **Make documents interactive**: Easily explore the content of your documents by asking questions.
- **Quick retrieval**: With the text stored in a vector store, you can query the content efficiently.


---

### ✨ **Enjoy the App!** ✨  
Now, go ahead and chat with your documents! πŸ˜„

---