examples/agentic_rag.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Agentic RAG implemented using FloAI\n",
    "\n",
    "FloAI has just made implementing agentic RAG simple and easy to manage. The Diagram shows what we are going to implement:\n",
    "\n",
    "<div style=\"text-align: center;\">\n",
    "    <img src=\"./images/agentic-rag.png\" alt=\"Sample Image\" width=\"1000\" height=\"300\"/>\n",
    "</div>\n",
    "\n",
    "To implement this we use the following components, already available in flo-ai:\n",
    "\n",
    "1. `tool` agent: This will be a tool agent to retrieve the records from vector store\n",
    "2. `delegator` agent: The agent will check if the retrieved records are relevent, else it will re-write the query and send it back for re-retrieval\n",
    "3. `llm` agent: This will generate the output from the relevant documents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from flo_ai import Flo\n",
    "from flo_ai import FloSession\n",
    "from langchain_openai import ChatOpenAI, OpenAIEmbeddings\n",
    "from langchain_chroma import Chroma\n",
    "from langchain_community.document_loaders import TextLoader\n",
    "from langchain_community.embeddings.sentence_transformer import (\n",
    "    SentenceTransformerEmbeddings,\n",
    ")\n",
    "from langchain_text_splitters import CharacterTextSplitter\n",
    "\n",
    "from dotenv import load_dotenv\n",
    "load_dotenv()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up Vector Store with Sample Data\n",
    "\n",
    "Using Chroma is this example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/vizsatiz/Documents/hub/flo/.venv/lib/python3.11/site-packages/sentence_transformers/cross_encoder/CrossEncoder.py:11: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from tqdm.autonotebook import tqdm, trange\n",
      "/Users/vizsatiz/Documents/hub/flo/.venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py:1601: FutureWarning: `clean_up_tokenization_spaces` was not set. It will be set to `True` by default. This behavior will be depracted in transformers v4.45, and will be then set to `False` by default. For more details check this issue: https://github.com/huggingface/transformers/issues/31884\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "# load the document and split it into chunks\n",
    "loader = TextLoader(\"./data/rag_document.txt\")\n",
    "documents = loader.load()\n",
    "\n",
    "# split it into chunks\n",
    "text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)\n",
    "docs = text_splitter.split_documents(documents)\n",
    "\n",
    "# create the open-source embedding function\n",
    "embedding_function = SentenceTransformerEmbeddings(model_name=\"all-MiniLM-L6-v2\")\n",
    "\n",
    "# load it into Chroma\n",
    "db = Chroma.from_documents(docs, embedding_function)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating the retrival tool\n",
    "\n",
    "This tool will retrive the records from vector db"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-09-23 14:42:29,261 - SESSION - INFO - New FloSession created with ID: fb386a6f-07b9-4cf4-a83a-21f322c7289c\n",
      "2024-09-23 14:42:29,359 - SESSION - INFO - Tool 'HousingLoanTool' registered for session fb386a6f-07b9-4cf4-a83a-21f322c7289c\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<flo_ai.state.flo_session.FloSession at 0x15b399450>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from flo_ai.retrievers.flo_retriever import FloRagBuilder\n",
    "from flo_ai.retrievers.flo_compression_pipeline import FloCompressionPipeline\n",
    "\n",
    "llm = ChatOpenAI(temperature=0, model_name='gpt-4o-mini')\n",
    "session = FloSession(llm)\n",
    "builder = FloRagBuilder(session, db.as_retriever())\n",
    "compression_pipeline = FloCompressionPipeline(OpenAIEmbeddings(model=\"text-embedding-3-small\"))\n",
    "compression_pipeline.add_embedding_reduntant_filter()\n",
    "compression_pipeline.add_embedding_relevant_filter()\n",
    "\n",
    "retriever_tool = builder.with_compression(compression_pipeline).build_retriever_tool(name=\"HousingLoanRetreiver\",\n",
    "                                                                                               description=\"Tool to fetch data around housing loans\")\n",
    "session.register_tool(name=\"HousingLoanTool\", tool=retriever_tool)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating the flo\n",
    "\n",
    "This follow as you can see consist of the following components:\n",
    "\n",
    "1. HousingLoanRetriver: This is a `tool agent`, with ability to retrieve from vector store\n",
    "2. RelevancyChecker: This is a `delegator agent` with ability to delegate based on relevancy\n",
    "3. ResponseGenerator: This is a `llm agent` with ability to response based on documents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-09-23 14:45:06,389 - BUILDER - INFO - Building Flo instance from YAML\n",
      "2024-09-23 14:45:06,398 - COMMON - INFO - Flo instance created for session fb386a6f-07b9-4cf4-a83a-21f322c7289c\n"
     ]
    },
    {
     "data": {
      "image/jpeg": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "agent_yaml = \"\"\"\n",
    "apiVersion: flo/alpha-v1\n",
    "kind: FloRoutedTeam\n",
    "name: agentic-rag\n",
    "team:\n",
    "    name: AgenticRAGTeam\n",
    "    router:\n",
    "        name: SupportSupervisor\n",
    "        kind: linear\n",
    "    agents:\n",
    "      - name: HousingLoanRetriver\n",
    "        kind: tool\n",
    "        job: Fetching the loan information from the loan tool\n",
    "        tools:\n",
    "          - name: HousingLoanTool\n",
    "      - name: RelevancyChecker\n",
    "        kind: delegator\n",
    "        to:\n",
    "          - name: HousingLoanRetriver\n",
    "          - name: ResponseGenerator\n",
    "        retry: 1\n",
    "        job: >\n",
    "          Your job is to check if the records fetched by the retriever are relevent to the question.\n",
    "          If its not relevant re-write the query send to HousingLoanRetriver for re-retrieval.\n",
    "          If its relevant then send to ResponseGenerator for response generation\n",
    "      - name: ResponseGenerator\n",
    "        kind: llm\n",
    "        job: Based on the documents given answer the user question that was asked\n",
    "\"\"\"\n",
    "\n",
    "flo = Flo.build(session, agent_yaml)\n",
    "\n",
    "flo.draw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-09-23 14:45:11,703 - COMMON - INFO - Invoking query for session fb386a6f-07b9-4cf4-a83a-21f322c7289c: Whats the interest rate for housing loan ?\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'messages': [HumanMessage(content='Whats the interest rate for housing loan ?'), HumanMessage(content='You can use any one of the ways to repay the loan:\\n \\n  Cheques\\n  Standing instructions at your branch\\n  FedNet - Internet Banking\\n  Automated Payment through ECS\\n  Mobile Banking\\n\\n  Housing loan interest rates will change subject to the changes made by Bank/RBI from time to time. \\n\\nPresent Repo Rate\\t 6.50%  (p.a)\\n \\n\\nLoan scheme\\n\\nInterest Rate (%) * \\n\\nHome Loan  \\n\\n8.80 (Repo Rate+ 2.30) Onwards \\n\\n*T&C Apply\\nHousing Loan\\n\\nYour dream home is never far away! Get hassle free home loans from Federal Bank to turn your dream home into reality. We assist you to realize your dream home. Avail your Housing Loan from us at competitive interest rates. The loan scheme assists borrowers for construction of house, acquisition of land & construction of house, repairs / renovation / remodeling / extension of house, reimbursement of debt incurred for construction / purchase / furnishing / beautification / purchase of flat / villa / house plots / takeover of housing loans / supplementary housing loan to employees of well-run companies / purchase of house plot for subsequent construction of house etc.', name='HousingLoanRetriver'), HumanMessage(content='The interest rate for housing loans currently starts at 8.80% per annum, which is based on the present repo rate of 6.50% plus an additional margin of 2.30%. Please note that interest rates may change based on adjustments made by the bank or the Reserve Bank of India (RBI) over time.', name='ResponseGenerator')], 'next': 'ResponseGenerator', 'loop_tracker': {'RelevancyChecker': 1}}\n"
     ]
    }
   ],
   "source": [
    "print(flo.invoke(\"Whats the interest rate for housing loan ?\"))"
   ]
  }
 ],
 "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}