Upload 4 files

app.py

@@ -0,0 +1,139 @@
+from fastapi import FastAPI, UploadFile, File, HTTPException
+from tools import update_retriever
+from chatbot import app as app_graph, rebuild_graph
+from langchain_core.messages import HumanMessage
+import os
+from fastapi.responses import StreamingResponse, FileResponse
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+from utils import TTS, STT
+
+# =====================================================
+# APP SETUP
+# =====================================================
+
+app = FastAPI()
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+# =====================================================
+# MODELS
+# =====================================================
+
+class TTSRequest(BaseModel):
+    text: str
+
+
+UPLOAD_DIR = "uploads"
+
+# =====================================================
+# HEALTH CHECK
+# =====================================================
+
+@app.get("/")
+def health():
+    return {"Status": "The api is live and running"}
+
+
+# =====================================================
+# FILE UPLOAD (RAG)
+# =====================================================
+
+@app.post("/upload")
+async def upload_file(file: UploadFile = File(...)):
+    os.makedirs(UPLOAD_DIR, exist_ok=True)
+
+    file_path = os.path.join(UPLOAD_DIR, file.filename)
+
+    with open(file_path, "wb") as f:
+        f.write(await file.read())
+
+    # Update retriever
+    update_retriever(file_path)
+
+    # 🔥 Rebuild LangGraph so RAG is active
+    rebuild_graph()
+
+    return {
+        "status": "success",
+        "filename": file.filename
+    }
+
+
+# =====================================================
+# CHAT ENDPOINT (STREAMING)
+# =====================================================
+
+@app.post("/chat")
+async def chat(message: str, session_id: str = "default"):
+
+    async def event_generator():
+        async for chunk in app_graph.astream(
+            {"messages": [HumanMessage(content=message)]},
+            config={"configurable": {"thread_id": session_id}},
+            stream_mode="messages"
+        ):
+            if not chunk:
+                continue
+
+            msg = chunk[0] if isinstance(chunk, tuple) else chunk
+
+            if hasattr(msg, "content") and msg.content:
+                cleaned = msg.content.strip()
+                if cleaned:
+                    yield f"data: {cleaned}\n\n"
+
+
+    return StreamingResponse(
+        event_generator(),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+            "Access-Control-Allow-Origin": "*",
+        },
+    )
+
+
+# =====================================================
+# STT
+# =====================================================
+
+@app.post("/stt")
+async def transcribe_audio(file: UploadFile = File(...)):
+    try:
+        return await STT(file)
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+# =====================================================
+# TTS
+# =====================================================
+
+@app.post("/tts")
+async def generate_tts(request: TTSRequest):
+    try:
+        if not request.text.strip():
+            raise HTTPException(status_code=400, detail="Text is empty")
+
+        audio_path = await TTS(text=request.text)
+
+        if not os.path.exists(audio_path):
+            raise HTTPException(status_code=500, detail="Audio file not created")
+
+        return FileResponse(
+            path=audio_path,
+            media_type="audio/mpeg",
+            filename="speech.mp3"
+        )
+
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))

chatbot.py

@@ -0,0 +1,115 @@
+from typing import TypedDict, Annotated
+from langchain_core.messages import BaseMessage, HumanMessage, SystemMessage
+from langgraph.checkpoint.memory import MemorySaver
+from tools import (
+    create_rag_tool,
+    arxiv_search,
+    calculator,
+    get_stock_price,
+    wikipedia_search,
+    tavily_search,
+    convert_currency,
+    unit_converter,
+    get_news,
+    get_joke,
+    get_quote,
+    get_weather,
+)
+from langchain_openai import ChatOpenAI
+from langgraph.graph import StateGraph, START
+from langgraph.graph.message import add_messages
+from langgraph.prebuilt import ToolNode, tools_condition
+from dotenv import load_dotenv
+import os
+
+load_dotenv()
+
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+
+# =====================================================
+# SYSTEM PROMPT
+# =====================================================
+
+SYSTEM_PROMPT = SystemMessage(
+    content="""
+You are an intelligent AI assistant built inside a LangGraph-based system created by Junaid.
+
+You MUST use RAG when a document has been uploaded.
+If no document contains the answer, say so clearly.
+Never hallucinate document content.
+"""
+)
+
+# =====================================================
+# STATE
+# =====================================================
+
+class ChatState(TypedDict):
+    messages: Annotated[list[BaseMessage], add_messages]
+
+
+# =====================================================
+# LLM
+# =====================================================
+
+llm = ChatOpenAI(
+    model="gpt-4.1-nano",
+    temperature=0.4,
+    streaming=True
+)
+
+
+# =====================================================
+# GRAPH BUILDER (🔥 IMPORTANT)
+# =====================================================
+
+memory = MemorySaver()
+app = None
+
+
+def build_graph():
+    global app
+
+    rag_tool = create_rag_tool()
+
+    tools = [
+        rag_tool,
+        get_stock_price,
+        calculator,
+        wikipedia_search,
+        arxiv_search,
+        tavily_search,
+        convert_currency,
+        unit_converter,
+        get_news,
+        get_joke,
+        get_quote,
+        get_weather,
+    ]
+
+    llm_with_tools = llm.bind_tools(tools)
+    tool_node = ToolNode(tools)
+
+    def chatbot(state: ChatState):
+        messages = [SYSTEM_PROMPT] + state["messages"]
+        response = llm_with_tools.invoke(messages)
+        return {"messages": [response]}
+
+    graph = StateGraph(ChatState)
+
+    graph.add_node("chat", chatbot)
+    graph.add_node("tools", tool_node)
+
+    graph.add_edge(START, "chat")
+    graph.add_conditional_edges("chat", tools_condition)
+    graph.add_edge("tools", "chat")
+
+    app = graph.compile(checkpointer=memory)
+
+
+# initial build
+build_graph()
+
+
+def rebuild_graph():
+    build_graph()

tools.py

@@ -0,0 +1,282 @@
+from langchain_core.tools import tool
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import FAISS
+from langchain_community.document_loaders import PyPDFLoader
+from langchain_openai import OpenAIEmbeddings
+from langchain_community.tools import WikipediaQueryRun, ArxivQueryRun
+from langchain_community.utilities import WikipediaAPIWrapper, ArxivAPIWrapper
+from langchain_core.tools import tool
+from langchain_community.tools.tavily_search import TavilySearchResults
+from dotenv import load_dotenv
+import os
+import requests
+
+load_dotenv()
+
+API_KEY = os.getenv("ALPHAVANTAGE_API_KEY")
+NEWS_API_KEY = os.getenv("NEWS_API_KEY")
+WEATHER_API_KEY = os.getenv("WEATHER_API_KEY")
+NEWS_API_KEY = os.getenv("NEWS_API_KEY")
+
+# -------------------------------
+# GLOBAL RETRIEVER
+# -------------------------------
+retriever = None
+
+
+def build_vectorstore(path: str):
+    loader = PyPDFLoader(path)
+    docs = loader.load()
+
+    splitter = RecursiveCharacterTextSplitter(
+        chunk_size=500,
+        chunk_overlap=100
+    )
+
+    split_docs = splitter.split_documents(docs)
+
+    embeddings = OpenAIEmbeddings(model="text-embedding-3-small")
+    return FAISS.from_documents(split_docs, embeddings)
+
+
+def update_retriever(pdf_path: str):
+    global retriever
+    vectorstore = build_vectorstore(pdf_path)
+    retriever = vectorstore.as_retriever(search_kwargs={"k": 3})
+
+
+def create_rag_tool():
+
+    @tool
+    def rag_search(query: str) -> str:
+        """
+        Retrieve relevant information from the uploaded document.
+
+        Use this tool when the user asks questions related to the uploaded PDF
+        or any document-based knowledge. If no document is available or no
+        relevant information is found, return an appropriate message.
+        """
+        global retriever
+
+        if retriever is None:
+            return "No document uploaded yet."
+
+        docs = retriever.invoke(query)
+
+        if not docs:
+            return "No relevant information found in the uploaded document."
+
+        context = "\n\n".join(d.page_content for d in docs)
+
+        return f"""
+            You are given extracted content from a document.
+
+            Your task:
+            - Summarize the content clearly
+            - Use bullet points where appropriate
+            - Keep formatting clean and readable
+            - Do NOT repeat unnecessary text
+            - Do NOT mention that this came from a document
+
+            DOCUMENT CONTENT:
+            {context}
+            """
+
+
+    return rag_search
+
+
+
+@tool
+def arxiv_search(query: str) -> dict:
+    """
+    Search arXiv for academic papers related to the query.
+    """
+    try:
+        arxiv = ArxivQueryRun(api_wrapper=ArxivAPIWrapper())
+        results = arxiv.run(query)
+        return {"query": query, "results": results}
+    except Exception as e:
+        return {"error": str(e)}
+    
+@tool
+def calculator(first_num: float, second_num: float, operation: str) -> dict:
+    """
+    Perform a basic arithmetic operation on two numbers.
+    Supported operations: add, sub, mul, div
+    """
+    try:
+        if operation == "add":
+            result = first_num + second_num
+        elif operation == "sub":
+            result = first_num - second_num
+        elif operation == "mul":
+            result = first_num * second_num
+        elif operation == "div":
+            if second_num == 0:
+                return {"error": "Division by zero is not allowed"}
+            result = first_num / second_num
+        else:
+            return {"error": f"Unsupported operation '{operation}'"}
+        
+        return {"first_num": first_num, "second_num": second_num, "operation": operation, "result": result}
+    except Exception as e:
+        return {"error": str(e)}
+@tool
+def tavily_search(query: str) -> dict:
+    """
+    Perform a web search using Tavily,
+    also use it to get weather information,
+    Returns up to 5 search results.
+    """
+    try:
+        search = TavilySearchResults(max_results=5)
+        results = search.run(query)
+        return {"query": query, "results": results}
+    except Exception as e:
+        return {"error": str(e)}
+
+
+@tool
+def get_stock_price(symbol: str) -> dict:
+    """
+    Fetch latest stock price for a given symbol (e.g. 'AAPL', 'TSLA') 
+    using Alpha Vantage with API key in the URL.
+    """
+    url = f"https://www.alphavantage.co/query?function=GLOBAL_QUOTE&symbol={symbol}&apikey={API_KEY}"
+    r = requests.get(url)
+    return r.json()
+
+@tool
+def wikipedia_search(query: str) -> dict:
+    """
+    Search Wikipedia for a given query and return results.
+    """
+    try:
+        wiki = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper())
+        results = wiki.run(query)
+        return {"query": query, "results": results}
+    except Exception as e:
+        return {"error": str(e)}
+
+@tool
+def convert_currency(amount: float, from_currency: str, to_currency: str) -> dict:
+    """
+    Convert amount from one currency to another using Frankfurter API.
+    Example: convert_currency(100, "USD", "EUR")
+    """
+    try:
+        url = f"https://api.frankfurter.app/latest?amount={amount}&from={from_currency}&to={to_currency}"
+        r = requests.get(url)
+        return r.json()
+    except Exception as e:
+        return {"error": str(e)}
+@tool
+
+
+def unit_converter(value: float, from_unit: str, to_unit: str) -> dict:
+    """
+    Convert between metric/imperial units (supports: km<->miles, kg<->lbs, C<->F).
+    Example: unit_converter(10, "km", "miles")
+    """
+    try:
+        conversions = {
+            ("km", "miles"): lambda x: x * 0.621371,
+            ("miles", "km"): lambda x: x / 0.621371,
+            ("kg", "lbs"): lambda x: x * 2.20462,
+            ("lbs", "kg"): lambda x: x / 2.20462,
+            ("C", "F"): lambda x: (x * 9/5) + 32,
+            ("F", "C"): lambda x: (x - 32) * 5/9
+        }
+        if (from_unit, to_unit) not in conversions:
+            return {"error": f"Unsupported conversion: {from_unit} -> {to_unit}"}
+        result = conversions[(from_unit, to_unit)](value)
+        return {"value": value, "from": from_unit, "to": to_unit, "result": result}
+    except Exception as e:
+        return {"error": str(e)}
+
+
+
+@tool
+def get_news(query: str) -> dict:
+    """
+    Fetch latest news headlines for a given query.
+    Example: get_news("artificial intelligence")
+    """
+    try:
+        url = f"https://newsapi.org/v2/everything?q={query}&apiKey={NEWS_API_KEY}&language=en"
+        r = requests.get(url)
+        return r.json()
+    except Exception as e:
+        return {"error": str(e)}
+
+
+@tool
+def get_joke(category: str = "Any") -> dict:
+    """
+    Get a random joke. Categories: Programming, Misc, Pun, Spooky, Christmas, Any
+    Example: get_joke("Programming")
+    """
+    try:
+        url = f"https://v2.jokeapi.dev/joke/{category}"
+        r = requests.get(url)
+        return r.json()
+    except Exception as e:
+        return {"error": str(e)}
+
+@tool
+def get_quote(tag: str = "") -> dict:
+    """
+    Fetch a random quote. Optionally filter by tag (e.g., 'inspirational', 'technology').
+    Example: get_quote("inspirational")
+    """
+    try:
+        url = f"https://api.quotable.io/random"
+        if tag:
+            url += f"?tags={tag}"
+        r = requests.get(url)
+        return r.json()
+    except Exception as e:
+        return {"error": str(e)}
+
+@tool
+def get_weather(city: str) -> dict:
+    """
+    Get current weather for a given city using WeatherAPI.com.
+    Example: get_weather("London")
+    """
+    try:
+        url = f"http://api.weatherapi.com/v1/current.json?key={WEATHER_API_KEY}&q={city}&aqi=no"
+        r = requests.get(url)
+        data = r.json()
+
+        if "error" in data:
+            return {"error": data["error"]["message"]}
+
+        return {
+            "location": data["location"]["name"],
+            "country": data["location"]["country"],
+            "temperature_c": data["current"]["temp_c"],
+            "temperature_f": data["current"]["temp_f"],
+            "condition": data["current"]["condition"]["text"],
+            "humidity": data["current"]["humidity"],
+            "wind_kph": data["current"]["wind_kph"],
+            "wind_dir": data["current"]["wind_dir"]
+        }
+    except Exception as e:
+        return {"error": str(e)}
+    
+
+
+@tool
+def get_news(query: str) -> dict:
+    """
+    Fetch latest news headlines for a given query.
+    Example: get_news("artificial intelligence")
+    """
+    try:
+        url = f"https://newsapi.org/v2/everything?q={query}&apiKey={NEWS_API_KEY}&language=en"
+        r = requests.get(url)
+        return r.json()
+    except Exception as e:
+        return {"error": str(e)}

utils.py

@@ -0,0 +1,66 @@
+import os
+from uuid import uuid4
+import edge_tts
+from groq import Groq
+from dotenv import load_dotenv
+
+load_dotenv()
+
+client = Groq()
+
+# ==================================================
+# 🎙️ TEXT TO SPEECH (FIXED VOICE)
+# ==================================================
+
+DEFAULT_VOICE = "en-US-MichelleNeural"
+
+async def TTS(
+    text: str,
+    output_dir: str = "tts_outputs",
+    rate: str = "+0%",
+    pitch: str = "+0Hz"
+) -> str:
+
+    if not text.strip():
+        raise ValueError("Empty text")
+
+    os.makedirs(output_dir, exist_ok=True)
+
+    filename = f"{uuid4().hex}.mp3"
+    output_path = os.path.join(output_dir, filename)
+
+    communicate = edge_tts.Communicate(
+        text=text,
+        voice=DEFAULT_VOICE,
+        rate=rate,
+        pitch=pitch
+    )
+
+    await communicate.save(output_path)
+    return output_path
+
+
+# ==================================================
+# 🎧 SPEECH TO TEXT
+# ==================================================
+
+async def STT(audio_file):
+    os.makedirs("uploads", exist_ok=True)
+    file_path = f"uploads/{uuid4().hex}.wav"
+
+    with open(file_path, "wb") as f:
+        f.write(await audio_file.read())
+
+    with open(file_path, "rb") as f:
+        transcription = client.audio.transcriptions.create(
+            file=f,
+            model="whisper-large-v3-turbo",
+            response_format="verbose_json",
+            temperature=0.0
+        )
+
+    return {
+        "text": transcription.text,
+        "segments": transcription.segments,
+        "language": transcription.language
+    }