Added simple RAG component to web app

Dockerfile

@@ -13,6 +13,8 @@ FROM ghcr.io/astral-sh/uv:python3.12-bookworm-slim
 
 WORKDIR /app
 
+RUN mkdir -p /app/static/data
+
 # Create a non-root user
 RUN useradd -m -u 1000 user
 RUN chown -R user:user /app

backend/app/problem_generator.py

@@ -1,24 +1,61 @@
 from typing import List
+import json
 
-# from backend.app.vectorstore import get_vector_db
-from langchain.agents import AgentExecutor, create_openai_functions_agent
-from langchain_core.messages import AIMessage, BaseMessage, HumanMessage
-from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
-from langchain.output_parsers.openai_functions import JsonOutputFunctionsParser
+from langchain_core.prompts import ChatPromptTemplate
 from langchain_openai import ChatOpenAI
-
+from langchain_core.runnables import RunnablePassthrough
+from langchain_core.output_parsers import StrOutputParser
+from langchain.output_parsers.openai_functions import JsonOutputFunctionsParser
+from backend.app.vectorstore import get_vector_db
 
 class ProblemGenerator:
+    def __init__(self):
+        # Initialize prompts
+        self.system_role_prompt = """
+            You are a helpful assistant that generates questions based on a given context.
+        """
+
+        self.user_role_prompt = """
+            Based on the following context about {query}, generate 5 relevant and specific questions.
+            Make sure the questions can be answered using only the provided context.
+
+            Context: {context}
+
+            Generate 5 questions that test understanding of the material in the context.
+            
+            Return only a json object with the following format:
+            {{
+                "questions": ["question1", "question2", "question3", "question4", "question5"]
+            }}
+        """
+
+        # Initialize chain components
+        self.chat_prompt = ChatPromptTemplate.from_messages([
+            ("system", self.system_role_prompt),
+            ("user", self.user_role_prompt)
+        ])
+        
+        self.llm = ChatOpenAI(model="gpt-3.5-turbo", temperature=0.7)
+        self.retriever = get_vector_db().as_retriever(search_kwargs={"k": 2})
+        
+        # Build the RAG chain
+        self.rag_chain = (
+            {"context": self.retriever, "query": RunnablePassthrough()}
+            | self.chat_prompt
+            | self.llm
+            | StrOutputParser()
+        )
+
     def generate_problems(self, query: str) -> List[str]:
         """
-        Generate problems based on the user's query.
+        Generate problems based on the user's query using RAG.
+        
+        Args:
+            query (str): The topic to generate questions about
+            
+        Returns:
+            List[str]: A list of generated questions
         """
-        # For MVP, returning random sample questions
-        sample_questions = [
-            "What is the main purpose of this framework?",
-            "How do you install this tool?",
-            "What are the key components?",
-            "Explain the basic workflow",
-            "What are the best practices?"
-        ]
-        return sample_questions
+        raw_result = self.rag_chain.invoke(query)
+        result = json.loads(raw_result)
+        return result["questions"]

backend/tests/test_api.py

@@ -14,7 +14,7 @@ def test_crawl_endpoint():
 def test_problems_endpoint():
     response = client.post(
         "/api/problems/",
-        json={"user_query": "test query"}
+        json={"user_query": "RAG"}
     )
     assert response.status_code == 200
     assert "Problems" in response.json()

test_problem_gen_rag.ipynb

@@ -0,0 +1,153 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "[nltk_data] Downloading package punkt_tab to\n",
+      "[nltk_data]     /Users/ryanrodriguez/nltk_data...\n",
+      "[nltk_data]   Package punkt_tab is already up-to-date!\n",
+      "[nltk_data] Downloading package averaged_perceptron_tagger_eng to\n",
+      "[nltk_data]     /Users/ryanrodriguez/nltk_data...\n",
+      "[nltk_data]   Package averaged_perceptron_tagger_eng is already up-to-\n",
+      "[nltk_data]       date!\n"
+     ]
+    }
+   ],
+   "source": [
+    "from langchain.output_parsers.openai_functions import JsonOutputFunctionsParser\n",
+    "from langchain_core.prompts import ChatPromptTemplate\n",
+    "from langchain_openai import ChatOpenAI\n",
+    "from langchain.chains import create_retrieval_chain\n",
+    "from langchain.chains.combine_documents import create_stuff_documents_chain\n",
+    "from backend.app.vectorstore import get_vector_db"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "system_role_prompt = \"\"\"\n",
+    "    You are a helpful assistant that generates questions based on a given context.\n",
+    "\"\"\"\n",
+    "\n",
+    "user_role_prompt = \"\"\"\n",
+    "    Based on the following context about {query}, generate 5 relevant and specific questions.\n",
+    "    Make sure the questions can be answered using only the provided context.\n",
+    "\n",
+    "    Context: {context}\n",
+    "\n",
+    "    Generate 5 questions that test understanding of the material in the context.\n",
+    "    \n",
+    "    Return only a json object with the following format:\n",
+    "    {{\n",
+    "        \"questions\": [\"question1\", \"question2\", \"question3\", \"question4\", \"question5\"]\n",
+    "    }}\n",
+    "\"\"\"\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "chat_prompt = ChatPromptTemplate.from_messages([\n",
+    "    (\"system\", system_role_prompt),\n",
+    "    (\"user\", user_role_prompt)\n",
+    "])\n",
+    "\n",
+    "openai_chat_model = ChatOpenAI(model=\"gpt-3.5-turbo\", temperature=0.7)\n",
+    "\n",
+    "retriever = get_vector_db().as_retriever(search_kwargs={\"k\": 2})\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from langchain_core.runnables import RunnablePassthrough\n",
+    "from langchain_core.output_parsers import StrOutputParser\n",
+    "\n",
+    "simple_rag  = (\n",
+    "    {\"context\": retriever, \"query\": RunnablePassthrough(), \"num_questions\": RunnablePassthrough()}\n",
+    "    | chat_prompt\n",
+    "    | openai_chat_model\n",
+    "    | StrOutputParser()\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "raw_result = simple_rag.invoke(\"RAG\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['What are the two main components of a typical RAG application?',\n",
+       " 'What is the purpose of the indexing component in a RAG application?',\n",
+       " \"What are the steps involved in the 'Load' phase of indexing?\",\n",
+       " 'Why is splitting text into smaller chunks important in the context of RAG applications?',\n",
+       " 'How does the retrieval and generation component of a RAG application process user queries?']"
+      ]
+     },
+     "execution_count": 23,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import json\n",
+    "result = json.loads(raw_result)\n",
+    "result[\"questions\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.12.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}