{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Using Opik with OpenAI Agents\n",
    "\n",
    "Opik integrates with OpenAI Agents to provide a simple way to log traces and analyse for all OpenAI LLM calls. This works for all OpenAI models, including if you are using the streaming API.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating an account on Comet.com\n",
    "\n",
    "[Comet](https://www.comet.com/site?from=llm&utm_source=opik&utm_medium=colab&utm_content=openai&utm_campaign=opik) provides a hosted version of the Opik platform, [simply create an account](https://www.comet.com/signup?from=llm&utm_source=opik&utm_medium=colab&utm_content=openai&utm_campaign=opik) and grab your API Key.\n",
    "\n",
    "> You can also run the Opik platform locally, see the [installation guide](https://www.comet.com/docs/opik/self-host/overview/?from=llm&utm_source=opik&utm_medium=colab&utm_content=openai&utm_campaign=opik) for more information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%pip install --upgrade opik openai-agents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import opik\n",
    "\n",
    "opik.configure(use_local=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparing our environment\n",
    "\n",
    "First, we will set up our OpenAI API keys."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import getpass\n",
    "\n",
    "if \"OPENAI_API_KEY\" not in os.environ:\n",
    "    os.environ[\"OPENAI_API_KEY\"] = getpass.getpass(\"Enter your OpenAI API key: \")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logging traces\n",
    "\n",
    "In order to log traces to Opik, we need to wrap our OpenAI calls with the `track_openai` function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from agents import Agent, Runner\n",
    "from agents import set_trace_processors\n",
    "from opik.integrations.openai.agents import OpikTracingProcessor\n",
    "\n",
    "os.environ[\"OPIK_PROJECT_NAME\"] = \"openai-agents-demo\"\n",
    "\n",
    "set_trace_processors(processors=[OpikTracingProcessor()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create and run your agent\n",
    "agent = Agent(\n",
    "    name=\"Creative Assistant\", \n",
    "    instructions=\"You are a creative writing assistant that helps users with poetry and creative content.\",\n",
    "    model=\"gpt-4o-mini\"\n",
    ")\n",
    "\n",
    "# Use async Runner.run() instead of run_sync() in Jupyter notebooks\n",
    "result = await Runner.run(agent, \"Write a haiku about recursion in programming.\")\n",
    "print(result.final_output)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using it with the `track` decorator\n",
    "\n",
    "If you have multiple steps in your LLM pipeline, you can use the `track` decorator to log the traces for each step. If OpenAI is called within one of these steps, the LLM call with be associated with that corresponding step:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from agents import Agent, Runner, function_tool\n",
    "from opik import track\n",
    "\n",
    "@function_tool\n",
    "def calculate_average(numbers: list[float]) -> float:\n",
    "    return sum(numbers) / len(numbers)\n",
    "\n",
    "@function_tool  \n",
    "def get_recommendation(topic: str, user_level: str) -> str:\n",
    "    recommendations = {\n",
    "        \"python\": {\n",
    "            \"beginner\": \"Start with Python.org's tutorial, then try Python Crash Course book. Practice with simple scripts and built-in functions.\",\n",
    "            \"intermediate\": \"Explore frameworks like Flask/Django, learn about decorators, context managers, and dive into Python's data structures.\",\n",
    "            \"advanced\": \"Study Python internals, contribute to open source, learn about metaclasses, and explore performance optimization.\"\n",
    "        },\n",
    "        \"machine learning\": {\n",
    "            \"beginner\": \"Start with Andrew Ng's Coursera course, learn basic statistics, and try scikit-learn with simple datasets.\",\n",
    "            \"intermediate\": \"Dive into deep learning with TensorFlow/PyTorch, study different algorithms, and work on real projects.\",\n",
    "            \"advanced\": \"Research latest papers, implement algorithms from scratch, and contribute to ML frameworks.\"\n",
    "        }\n",
    "    }\n",
    "    \n",
    "    topic_lower = topic.lower()\n",
    "    level_lower = user_level.lower()\n",
    "    \n",
    "    if topic_lower in recommendations and level_lower in recommendations[topic_lower]:\n",
    "        return recommendations[topic_lower][level_lower]\n",
    "    else:\n",
    "        return f\"For {topic} at {user_level} level: Focus on fundamentals, practice regularly, and build projects to apply your knowledge.\"\n",
    "\n",
    "def create_advanced_agent():\n",
    "    \"\"\"Create an advanced agent with tools and comprehensive instructions.\"\"\"\n",
    "    instructions = \"\"\"\n",
    "    You are an expert programming tutor and learning advisor. You have access to tools that help you:\n",
    "    1. Calculate averages for performance metrics, grades, or other numerical data\n",
    "    2. Provide personalized learning recommendations based on topics and user experience levels\n",
    "    \n",
    "    Your role:\n",
    "    - Help users learn programming concepts effectively\n",
    "    - Provide clear, beginner-friendly explanations when needed\n",
    "    - Use your tools when appropriate to give concrete help\n",
    "    - Offer structured learning paths and resources\n",
    "    - Be encouraging and supportive\n",
    "    \n",
    "    When users ask about:\n",
    "    - Programming languages: Use get_recommendation to provide tailored advice\n",
    "    - Performance or scores: Use calculate_average if numbers are involved\n",
    "    - Learning paths: Combine your knowledge with tool-based recommendations\n",
    "    \n",
    "    Always explain your reasoning and make your responses educational.\n",
    "    \"\"\"\n",
    "    \n",
    "    return Agent(\n",
    "        name=\"AdvancedProgrammingTutor\",\n",
    "        instructions=instructions,\n",
    "        model=\"gpt-4o-mini\",\n",
    "        tools=[calculate_average, get_recommendation]\n",
    "    )\n",
    "\n",
    "advanced_agent = create_advanced_agent()\n",
    "\n",
    "advanced_queries = [\n",
    "    \"I'm new to Python programming. Can you tell me about it?\",\n",
    "    \"I got these test scores: 85, 92, 78, 96, 88. What's my average and how am I doing?\",\n",
    "    \"I know some Python basics but want to learn machine learning. What should I do next?\",\n",
    "    \"Can you help me calculate the average of these response times: 1.2, 0.8, 1.5, 0.9, 1.1 seconds? And tell me if that's good performance?\"\n",
    "]\n",
    "\n",
    "for i, query in enumerate(advanced_queries, 1):\n",
    "    print(f\"\\n📝 Query {i}: {query}\")\n",
    "    result = await Runner.run(advanced_agent, query)\n",
    "    print(f\"🤖 Response: {result.final_output}\")\n",
    "    print(\"=\" * 80)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The trace can now be viewed in the UI:\n",
    "\n",
    "![OpenAI Integration](https://raw.githubusercontent.com/comet-ml/opik/main/apps/opik-documentation/documentation/static/img/cookbook/openai_agents_cookbook.png)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}