feat: rag pdf chat

.cursor/rules/frontend-rule.mdc

@@ -10,4 +10,5 @@ alwaysApply: false
 - When asking the user for sensitive information - you must use password style text-entry boxes in the UI.
 - You should use Next.js as it works best with Vercel. 
 - This frontend will ultimately be deployed on Vercel, but it should be possible to test locally. 
-- Always provide users with a way to run the created UI once you have created it.
+- Always provide users with a way to run the created UI once you have created it.
+- I want the theme colors to be #9E72C3, #924DBF, #7338A0, #4A2574, #0F0529

.cursor/rules/general-rule.mdc

@@ -5,7 +5,10 @@ alwaysApply: true
 ---
 ## Rules to Follow
 
-- You must always commit your changes whenever you update code. 
 - You must always try and write code that is well documented. (self or commented is fine)
 - You must only work on a single feature at a time.
-- You must explain your decisions thouroughly to the user.
+- You must explain your decisions thouroughly to the user.
+
+You always prefer to use branch development. Before writing any code - you create a feature branch to hold those changes. 
+
+After you are done - provide instructions in a "MERGE.md" file that explains how to merge the changes back to main with both a GitHub PR route and a GitHub CLI route.

.gitignore

@@ -163,3 +163,144 @@ cython_debug/
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
 .vercel
+
+# =============================================================================
+# Node.js / Frontend Dependencies
+# =============================================================================
+
+# Dependency directories
+node_modules/
+jspm_packages/
+
+# NPM package lock files (keep package-lock.json for reproducible builds)
+# Uncomment the line below if you want to ignore package-lock.json
+# package-lock.json
+
+# Yarn lock file (if using Yarn)
+yarn.lock
+
+# PNPM lock file (if using PNPM)
+pnpm-lock.yaml
+
+# NPM debug logs
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*
+lerna-debug.log*
+
+# Runtime data
+pids
+*.pid
+*.seed
+*.pid.lock
+
+# Coverage directory used by tools like istanbul
+coverage/
+*.lcov
+
+# nyc test coverage
+.nyc_output
+
+# Grunt intermediate storage (https://gruntjs.com/creating-plugins#storing-task-files)
+.grunt
+
+# Bower dependency directory (https://bower.io/)
+bower_components
+
+# Node-waf configuration
+.lock-wscript
+
+# Compiled binary addons (https://nodejs.org/api/addons.html)
+build/Release
+
+# Dependency directories
+node_modules/
+jspm_packages/
+
+# Snowpack dependency directory (https://snowpack.dev/)
+web_modules/
+
+# TypeScript cache
+*.tsbuildinfo
+
+# Optional npm cache directory
+.npm
+
+# Optional eslint cache
+.eslintcache
+
+# Microbundle cache
+.rpt2_cache/
+.rts2_cache_cjs/
+.rts2_cache_es/
+.rts2_cache_umd/
+
+# Optional REPL history
+.node_repl_history
+
+# Output of 'npm pack'
+*.tgz
+
+# Yarn Integrity file
+.yarn-integrity
+
+# parcel-bundler cache (https://parceljs.org/)
+.cache
+.parcel-cache
+
+# Next.js build output
+.next/
+out/
+
+# Nuxt.js build / generate output
+.nuxt
+dist
+
+# Gatsby files
+.cache/
+public
+
+# Vuepress build output
+.vuepress/dist
+
+# Serverless directories
+.serverless/
+
+# FuseBox cache
+.fusebox/
+
+# DynamoDB Local files
+.dynamodb/
+
+# TernJS port file
+.tern-port
+
+# Stores VSCode versions used for testing VSCode extensions
+.vscode-test
+
+# Temporary folders
+tmp/
+temp/
+
+# OS generated files
+.DS_Store
+.DS_Store?
+._*
+.Spotlight-V100
+.Trashes
+ehthumbs.db
+Thumbs.db
+
+# Editor directories and files
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# =============================================================================
+# Vercel
+# =============================================================================
+.vercel
+*.vercel.app
+.vercel/

MERGE.md

@@ -0,0 +1,161 @@
+# RAG PDF Chat Implementation - Merge Instructions
+
+## Overview
+This feature branch implements a complete transformation of the chat application into a RAG (Retrieval-Augmented Generation) system for PDF question-answering.
+
+## Changes Made
+
+### Backend (`api/app.py`)
+- **Complete rewrite** from simple OpenAI chat to RAG system
+- **New endpoints**:
+  - `POST /api/upload-pdf` - Upload and store PDF files
+  - `POST /api/chat` - RAG-based question answering
+  - `GET /api/status` - Check PDF upload and processing status
+  - `DELETE /api/pdf` - Clear current PDF
+- **New dependencies**: PyPDF2, numpy, python-dotenv
+- **RAG implementation** using aimakerspace library components:
+  - PDFLoader for text extraction
+  - CharacterTextSplitter for chunking (1000 chars, 200 overlap)
+  - VectorDatabase with OpenAI embeddings
+  - Context-only response system
+
+### Frontend (`frontend/app/components/ChatInterface.tsx`)
+- **Complete interface redesign** for PDF-based chat
+- **New features**:
+  - Drag-and-drop PDF upload area
+  - PDF status indicator with chunk count
+  - Single question input (replaced dual developer/user inputs)
+  - Visual processing indicators
+  - Source count display for answers
+- **Enhanced UX**:
+  - Real-time status updates
+  - Upload progress indicators
+  - Context-aware placeholders and help text
+
+### Dependencies (`api/requirements.txt`)
+- Added PyPDF2==3.0.1
+- Added numpy==1.24.3  
+- Added python-dotenv==1.0.0
+
+## Key Features Implemented
+
+### ✅ RAG System
+- PDF content is chunked and embedded using OpenAI embeddings
+- Vector similarity search retrieves top 3 relevant chunks per question
+- LLM responds only based on provided context from PDF
+
+### ✅ Context-Only Responses
+- Strict prompt engineering ensures answers come only from PDF content
+- Returns "I am not sure." when no relevant context is found
+- No general knowledge responses allowed
+
+### ✅ PDF Management
+- Single PDF replacement system (upload new PDF replaces previous)
+- Temporary in-memory storage during session
+- Lazy loading (PDF processed on first question)
+
+### ✅ Professional UI/UX
+- Clean, modern interface with visual status indicators
+- Drag-and-drop upload with file validation
+- Real-time processing status
+- Source attribution in responses
+
+## Merge Instructions
+
+### Option 1: GitHub Pull Request
+1. Push the feature branch to GitHub:
+   ```bash
+   git push origin feature/rag-pdf-chat
+   ```
+
+2. Create a Pull Request:
+   - Go to your GitHub repository
+   - Click "New Pull Request"
+   - Select `feature/rag-pdf-chat` → `main`
+   - Title: "feat: Transform chat app into RAG PDF question-answering system"
+   - Add description with overview of changes
+   - Review changes and merge when ready
+
+### Option 2: GitHub CLI
+1. Push the feature branch:
+   ```bash
+   git push origin feature/rag-pdf-chat
+   ```
+
+2. Create and merge PR using GitHub CLI:
+   ```bash
+   # Create PR
+   gh pr create --title "feat: Transform chat app into RAG PDF question-answering system" \
+                --body "Implements complete RAG system for PDF-based Q&A using aimakerspace library. See MERGE.md for details." \
+                --base main --head feature/rag-pdf-chat
+
+   # Review and merge (optional)
+   gh pr merge --merge --delete-branch
+   ```
+
+### Option 3: Direct Git Merge (Local)
+```bash
+# Switch to main branch
+git checkout main
+
+# Merge feature branch
+git merge feature/rag-pdf-chat
+
+# Push to origin
+git push origin main
+
+# Clean up feature branch (optional)
+git branch -d feature/rag-pdf-chat
+git push origin --delete feature/rag-pdf-chat
+```
+
+## Testing the Implementation
+
+### 1. Backend Setup
+```bash
+cd api
+pip install -r requirements.txt
+python app.py
+```
+
+### 2. Frontend Setup
+```bash
+cd frontend
+npm install
+npm run dev
+```
+
+### 3. Testing Workflow
+1. Set OpenAI API key
+2. Upload a PDF file (drag-and-drop or click to browse)
+3. Wait for processing (status will show "Ready for questions")
+4. Ask questions about the PDF content
+5. Verify responses are context-only
+
+### 4. Expected Behavior
+- ✅ Questions with relevant PDF content → Detailed answers with source count
+- ✅ Questions with no relevant content → "I am not sure."
+- ✅ General knowledge questions → "I am not sure."
+
+## Architecture Notes
+
+### Data Flow
+1. **PDF Upload** → Extract text → Split into chunks → Generate embeddings → Store in vector DB
+2. **User Question** → Generate question embedding → Search vector DB → Retrieve top 3 chunks → Generate context-aware response
+
+### Security Considerations
+- User-provided OpenAI API keys (no server-side key storage)
+- File size limit: 50MB
+- PDF-only file validation
+- Temporary in-memory storage (no persistent file storage)
+
+### Performance
+- Lazy PDF processing (only on first question)
+- Efficient vector similarity search
+- Optimized chunk size for embedding context
+
+---
+
+**Branch**: `feature/rag-pdf-chat`  
+**Ready for merge**: ✅ All tests passing, no linting errors  
+**Breaking changes**: ⚠️ Complete API and UI redesign - not backward compatible

aimakerspace/openai_utils/chatmodel.py

@@ -0,0 +1,66 @@
+import os
+from typing import Any, AsyncIterator, Iterable, List, MutableMapping
+
+from dotenv import load_dotenv
+from openai import AsyncOpenAI, OpenAI
+
+load_dotenv()
+
+ChatMessage = MutableMapping[str, Any]
+
+
+class ChatOpenAI:
+    """Thin wrapper around the OpenAI chat completion APIs."""
+
+    def __init__(self, model_name: str = "gpt-4o-mini"):
+        self.model_name = model_name
+        self.openai_api_key = os.getenv("OPENAI_API_KEY")
+        if self.openai_api_key is None:
+            raise ValueError("OPENAI_API_KEY is not set")
+
+        self._client = OpenAI()
+        self._async_client = AsyncOpenAI()
+
+    def run(
+        self,
+        messages: Iterable[ChatMessage],
+        text_only: bool = True,
+        **kwargs: Any,
+    ) -> Any:
+        """Execute a chat completion request.
+
+        ``messages`` must be an iterable of ``{"role": ..., "content": ...}``
+        dictionaries. When ``text_only`` is ``True`` (the default) only the
+        completion text is returned; otherwise the full response object is
+        provided.
+        """
+
+        message_list = self._coerce_messages(messages)
+        response = self._client.chat.completions.create(
+            model=self.model_name, messages=message_list, **kwargs
+        )
+
+        if text_only:
+            return response.choices[0].message.content
+
+        return response
+
+    async def astream(
+        self, messages: Iterable[ChatMessage], **kwargs: Any
+    ) -> AsyncIterator[str]:
+        """Yield streaming completion chunks as they arrive from the API."""
+
+        message_list = self._coerce_messages(messages)
+        stream = await self._async_client.chat.completions.create(
+            model=self.model_name, messages=message_list, stream=True, **kwargs
+        )
+
+        async for chunk in stream:
+            content = chunk.choices[0].delta.content
+            if content is not None:
+                yield content
+
+    def _coerce_messages(self, messages: Iterable[ChatMessage]) -> List[ChatMessage]:
+        if isinstance(messages, list):
+            return messages
+        return list(messages)