Travel_Planner.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Virtual Travel Planning Agent:\n",
    "## Done By:\n",
    "* Aakar Mutha : aakar.mutha@nyu.edu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from dotenv import load_dotenv\n",
    "\n",
    "_ = load_dotenv()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# General Imports\n",
    "from langgraph.graph import StateGraph, END\n",
    "from typing import TypedDict, Annotated, List\n",
    "from langgraph.checkpoint.sqlite import SqliteSaver\n",
    "from langchain_core.messages import AnyMessage, SystemMessage, HumanMessage, AIMessage, ChatMessage\n",
    "from langchain_core.pydantic_v1 import BaseModel\n",
    "\n",
    "from tavily import TavilyClient\n",
    "import os\n",
    "\n",
    "# LLM model import\n",
    "from langchain_anthropic import ChatAnthropic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Agent State Definition \n",
    "class AgentState(TypedDict):\n",
    "    task: str\n",
    "    plan: str\n",
    "    draft: str\n",
    "    critique: str\n",
    "    queries: List[str]\n",
    "    answers: List[str]\n",
    "    revision_number: int\n",
    "    max_revisions: int\n",
    "\n",
    "class Queries(BaseModel):\n",
    "    queries: List[str]\n",
    "    \n",
    "    \n",
    "## Configeration Variables\n",
    "memory = SqliteSaver.from_conn_string(\":memory:\")\n",
    "\n",
    "model = ChatAnthropic(model=\"claude-3-haiku-20240307\")\n",
    "\n",
    "tavily = TavilyClient(api_key=os.environ[\"TAVILY_API_KEY\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Defining Prompts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "VACATION_PLANNING_SUPERVISOR_PROMPT=\"\"\"You are the vacation planning supervisor. You have to give an outline of what the planning agent \\\n",
    "has to consider when planning the vacation according to the user input.\"\"\"\n",
    "\n",
    "PLANNER_ASSISTANT_PROMPT = \"\"\"You are an assistant charged with providing information that can be used by the planner to plan the vacation.\n",
    "Generate a list of search queries that will be useful for the planner. Generate a maximum of 3 queries.\"\"\"\n",
    "\n",
    "VACATION_PLANNER_PROMPT = \"\"\"You are an expert vacation planner tasked with suggesting vacation itineraries.\n",
    "You will provide the user with a suggestion of a vacation spot based on the outline and research.\n",
    "If the user provides tweeks or changes respond with updated versions of the itineraries.\n",
    "Always output a detailed daily itinerary in the following format even after integrating the suggestions:\n",
    "------\n",
    "Place:\n",
    "Dates:\n",
    "Estimated spending: xx USD\n",
    "Mode of commute from <Origin>: yy\n",
    "Estimated time to reach <Destination>: zzz Hrs vv Minutes\n",
    "Itenary:\n",
    "   Day 1: <DATE>\n",
    "   - Things the user can do.\n",
    "   - Things the user can do.\n",
    "   - Things the user can do.\n",
    "  \n",
    "   Day 2: <DATE>\n",
    "   - Things the user can do.\n",
    "   - Things the user can do.\n",
    "   - Things the user can do.\n",
    "  \n",
    "  \n",
    "------   \n",
    "Utilize the information below as needed:\n",
    "------\n",
    "{answers}\"\"\"\n",
    "\n",
    "PLANNER_CRITIQUE_PROMPT = \"\"\"Your duty is to criticize the planning done by the vacation planner.\n",
    "In your response include if you agree with options presented by the planner, if not then give detailed suggestions on what should be changed.\n",
    "You can also suggest some other destination that should be checked out.\n",
    "\"\"\"\n",
    "\n",
    "PLANNER_CRITIQUE_ASSISTANT_PROMPT = \"\"\"You are a assistant charged with providing information that can be used to make any requested revisions.\n",
    "Generate a list of search queries that will gather any relevent information. Only generate 3 queries max. \n",
    "You should consider the queries and answers that were previoulsy used:\n",
    "QUERIES:\n",
    "{queries}\n",
    "\n",
    "ANSWERS:\n",
    "{answers}\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def instruction_node(state: AgentState):\n",
    "#     messages = [\n",
    "#         SystemMessage(content=VACATION_PLANNING_INSTRCTION_PROMPT),\n",
    "#         HumanMessage(content=state['task'])\n",
    "#     ]\n",
    "#     task = model.invoke(messages)\n",
    "#     return {'task' : task.content}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plan_node(state: AgentState):\n",
    "    messages = [\n",
    "        SystemMessage(content=VACATION_PLANNING_SUPERVISOR_PROMPT), \n",
    "        HumanMessage(content=state['task'])\n",
    "    ]\n",
    "    response = model.invoke(messages)\n",
    "    print(\"**********************************************************\")\n",
    "    print(\"Plan: \")\n",
    "    print(response.content)\n",
    "    print(\"**********************************************************\")\n",
    "    return {\"plan\": response.content}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def research_plan_node(state: AgentState):\n",
    "    pastQueries = state['queries'] or []\n",
    "    answers = state['answers'] or []\n",
    "    queries = model.with_structured_output(Queries).invoke([\n",
    "        SystemMessage(content=PLANNER_ASSISTANT_PROMPT),\n",
    "        HumanMessage(content=state['plan'])\n",
    "    ])   \n",
    "    print(\"**********************************************************\")\n",
    "    print(\"Queries and Response: \")\n",
    "    for q in queries.queries:\n",
    "        print(\"Query: \" + q)\n",
    "        pastQueries.append(q)\n",
    "        response = tavily.search(query=q, max_results=2)\n",
    "        for r in response['results']:\n",
    "            print(\"Tavily Response: \" + r['content'])\n",
    "            answers.append(r['content'])\n",
    "    print(\"**********************************************************\")\n",
    "    return {\n",
    "        \"queries\": pastQueries,\n",
    "        \"answers\": answers\n",
    "        }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generation_node(state: AgentState):\n",
    "    answers = \"\\n\\n\".join(state['answers'] or [])\n",
    "    user_message = HumanMessage(\n",
    "        content=f\"{state['task']}\\n\\nHere is my plan:\\n\\n{state['plan']}\")\n",
    "    messages = [\n",
    "        SystemMessage(\n",
    "            content=VACATION_PLANNER_PROMPT.format(answers=answers)\n",
    "        ),\n",
    "        user_message\n",
    "        ]\n",
    "    response = model.invoke(messages)\n",
    "    print(\"**********************************************************\")\n",
    "    print(\"Draft: \")\n",
    "    print(response.content)\n",
    "    print(\"**********************************************************\")\n",
    "    return {\n",
    "        \"draft\": response.content, \n",
    "        \"revision_number\": state.get(\"revision_number\", 1) + 1\n",
    "    }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def reflection_node(state: AgentState):\n",
    "    messages = [\n",
    "        SystemMessage(content=PLANNER_CRITIQUE_PROMPT), \n",
    "        HumanMessage(content=state['draft'])\n",
    "    ]\n",
    "    response = model.invoke(messages)\n",
    "    print(\"**********************************************************\")\n",
    "    print(\"Critique: \")\n",
    "    print(response.content)\n",
    "    print(\"**********************************************************\")\n",
    "    return {\"critique\": response.content}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def research_critique_node(state: AgentState):\n",
    "    pastQueries = state['queries'] or []\n",
    "    answers = state['answers'] or []\n",
    "    queries = model.with_structured_output(Queries).invoke([\n",
    "        SystemMessage(content=PLANNER_CRITIQUE_ASSISTANT_PROMPT.format(queries=pastQueries, answers=answers)),\n",
    "        HumanMessage(content=state['critique'])\n",
    "    ])\n",
    "    print(\"**********************************************************\")\n",
    "    print(\"Queries and Response:\")\n",
    "    for q in queries.queries:\n",
    "        print(\"Query: \" + q)\n",
    "        pastQueries.append(q)\n",
    "        response = tavily.search(query=q, max_results=2)\n",
    "        for r in response['results']:\n",
    "            print(\"Tavily Response: \" + r['content'])\n",
    "            answers.append(r['content'])\n",
    "    print(\"**********************************************************\")       \n",
    "    return {\n",
    "        \"queries\": pastQueries,\n",
    "        \"answers\": answers\n",
    "        }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def should_continue(state):\n",
    "    if state[\"revision_number\"] > state[\"max_revisions\"]:\n",
    "        return END\n",
    "    return \"reflect\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "builder = StateGraph(AgentState)\n",
    "\n",
    "builder.add_node(\"planner\", plan_node)\n",
    "builder.add_node(\"research_plan\", research_plan_node)\n",
    "builder.add_node(\"generate\", generation_node)\n",
    "builder.add_node(\"reflect\", reflection_node)\n",
    "builder.add_node(\"research_critique\", research_critique_node)\n",
    "\n",
    "builder.set_entry_point(\"planner\")\n",
    "\n",
    "builder.add_conditional_edges(\n",
    "    \"generate\", \n",
    "    should_continue, \n",
    "    {END: END, \"reflect\": \"reflect\"}\n",
    ")\n",
    "\n",
    "builder.add_edge(\"planner\", \"research_plan\")\n",
    "builder.add_edge(\"research_plan\", \"generate\")\n",
    "\n",
    "builder.add_edge(\"reflect\", \"research_critique\")\n",
    "builder.add_edge(\"research_critique\", \"generate\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "graph = builder.compile(checkpointer=memory)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "\n",
    "Image(graph.get_graph().draw_png())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "**********************************************************\n",
      "Plan: \n",
      "Certainly! Here's an outline of what the planning agent should consider when planning a day trip for a group of 4 friends residing in NYC:\n",
      "\n",
      "I. Destination Selection\n",
      "   A. Proximity to NYC\n",
      "      1. Consider destinations within a 2-3 hour drive from NYC\n",
      "      2. Prioritize destinations that are easily accessible by public transportation or car\n",
      "   B. Diverse Interests\n",
      "      1. Cater to the varied interests of the group (e.g., nature, history, culture, food)\n",
      "      2. Ensure the destination offers a range of activities to accommodate everyone's preferences\n",
      "\n",
      "II. Transportation\n",
      "   A. Mode of Transportation\n",
      "      1. Evaluate the best mode of transportation (e.g., private car, train, bus)\n",
      "      2. Consider the group's comfort and convenience\n",
      "   B. Parking Availability\n",
      "      1. Assess the parking situation at the destination\n",
      "      2. Provide guidance on parking options and costs\n",
      "\n",
      "III. Itinerary Planning\n",
      "   A. Prioritize Must-See Attractions\n",
      "      1. Research and identify the top attractions or landmarks in the chosen destination\n",
      "      2. Ensure the group can comfortably experience the highlights\n",
      "   B. Allocate Time for Exploration\n",
      "      1. Plan for ample time to explore the destination at a leisurely pace\n",
      "      2. Factor in travel time between activities\n",
      "   C. Incorporate Breaks and Meals\n",
      "      1. Schedule breaks for refreshments and meals\n",
      "      2. Recommend restaurants or picnic spots that cater to the group's dietary preferences\n",
      "\n",
      "IV. Accommodation and Logistics\n",
      "   A. Lodging (if needed)\n",
      "      1. Suggest suitable accommodation options for a day trip, such as hotels or Airbnb rentals\n",
      "      2. Provide information on check-in and check-out times\n",
      "   B. Packing Recommendations\n",
      "      1. Advise the group on appropriate attire and essential items to bring\n",
      "      2. Provide tips on managing luggage and personal belongings\n",
      "\n",
      "V. Budget and Cost Estimation\n",
      "   A. Transportation Costs\n",
      "      1. Estimate the cost of the chosen mode of transportation\n",
      "      2. Include any parking or tolls\n",
      "   B. Admission Fees\n",
      "      1. Research and provide information on the admission fees for the selected attractions\n",
      "   C. Meal and Refreshment Costs\n",
      "      1. Suggest a budget for meals and refreshments during the day trip\n",
      "\n",
      "VI. Emergency Planning\n",
      "   A. Contact Information\n",
      "      1. Provide emergency contact numbers for the destination\n",
      "      2. Share the contact information of the planning agent\n",
      "   B. Backup Plan\n",
      "      1. Identify alternative options in case of unexpected events or weather conditions\n",
      "\n",
      "By considering these key elements, the planning agent can create a comprehensive and enjoyable day trip for the group of 4 friends residing in NYC.\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Queries and Response: \n",
      "Query: places to visit near new york city\n",
      "Tavily Response: The best wineries in NJ\n",
      "At these wineries in NJ, sample some of the Garden State’s finest vintages while you admire the green scenery\n",
      "Sip your way through four weekend getaways to nearby orchards\n",
      "Hit the road (with a designated driver) to taste the brews, wines and ciders of farms and orchards less than three hours away!\n",
      "Where to stay on weekend getaways\n",
      "The best Airbnb Hamptons rentals\n",
      "For summer getaway inspiration, look no further than these gorgeous Airbnb Hamptons rentals\n",
      "13 cozy cabins near NYC that you can rent on Airbnb\n",
      "Whether you need lodging near a ski resort or somewhere rustic chic for a girls’ weekend, this list of cabins near NYC on Airbnb has got you covered\n",
      "The best Hudson, NY, hotels and B&Bs\n",
      "The best Hudson, NY, hotels and bed-and-breakfasts for the next time you need to escape the city for a relaxing getaway\n",
      "The best bed-and-breakfasts in NJ\n",
      "Looking for a great bed-and-breakfast? Five dramatic lighthouses that you can rent overnight near NYC\n",
      "Not only will you gain some perspective during your stay, but you’ll literally be able to see for miles\n",
      "Three architecturally stunning geodesic domes you can rent right now\n",
      "These vacation rentals all offer the chance to be immersed in some pretty stunning architecture\n",
      "Weekend getaways: City breaks\n",
      "Atlantic City\n",
      "Chicago\n",
      "Los Angeles\n",
      "Boston\n",
      "Las Vegas\n",
      "Miami\n",
      "San Fancisco\n",
      "Toronto\n",
      "Archive coverage of weekend getaways\n",
      "Princeton, NJ (2008)\n",
      " Do this: The state park, situated in the mountains rising from the west bank of the Hudson River, which offers a bevy of hiking and biking trails, as well as picnic groves, lake and river fishing access, a swimming pool and a zoo.\n",
      "Stay here: Bear Mountain Inn, originally built in 1915, hosted Eleanor Roosevelt, Harry Truman and Dwight Eisenhower back in the day, and has been extensively renovated to include 15 luxury guest rooms, plus a spa and a restaurant.\n",
      " The best weekend getaways from NYC\n",
      "Escape the hustle and bustle with these weekend getaways from NYC\n",
      "While we can never quit NYC, with its amazing restaurants, internationally-recognized bars and world-class museums, the crowds, the hustle and the pace, however, can lead to some serious stress and, well, sometimes we need a weekend trip.\n",
      " 15 gorgeous Airbnbs for ski getaways from NYC\n",
      "Rent one of these log cabins, chic chalets, modern townhouses or other Airbnbs for ski getaways from NYC\n",
      "The best weekend trips from NYC for history buffs\n",
      "Discover the area’s rich history on these weekend trips to gorgeous mansions, intriguing museums and cool ships\n",
      "12 places to go this summer when everyone else is at Fire Island\n",
      "\n",
      "Tavily Response: Top day trips from NYC. 1. Woodstock, NY. Though the name conjures a crowd splashed in tie-dye and the faint scent of marijuana, Woodstock isn't actually where the 1969 festival was held. (That ...\n",
      "Query: best day trips from nyc\n",
      "Tavily Response: Find out the best places to escape from NYC for a day of nature, culture, history or adventure. Whether you want to hike, shop, eat, drink or relax, there's a destination for you within two hours of the city.\n",
      "Tavily Response: The park is roughly 12 miles long, covers 2,500 acres, and is a…\n",
      "07\n",
      "Day Trip from NYC for Nature Lovers: Cold Spring\n",
      "Cold Spring is renowned for being one of the best-preserved towns from the 19th Century and is just a 90-minute train ride from New York…\n",
      "08\n",
      "Day Trip from NYC for Nature Lovers: Mohonk Preserve\n",
      "Located in the town of New Paltz, New York, you’ll find a private 8,000-acre nature preserve that’s highly popular among tourists and residents alike. If you decide to take a day trip to Washington D.C., don’t forget to check out these sights:\n",
      "Things to do in Washington D.C. on your day trip from New York\n",
      "Places To See in Washington D.C.\n",
      "Helpful links\n",
      "If you looking for an accommodation check our this website and if you want to rent a car in New York, I recommend this site.\n",
      " If you decide to take a day trip to Boston, don’t forget to check out these sights:\n",
      "Things to do in Boston on your day trip from New York\n",
      "Places To See in Boston\n",
      "Helpful links\n",
      "If you looking for an accommodation check our this website and if you want to rent a car in New York, I recommend this site.\n",
      " If you decide to take a day trip to Philadelphia, don’t forget to check out these sights:\n",
      "Things to do in Philadelphia on your day trip from New York\n",
      "Places To See in Philadelphia\n",
      "Helpful links\n",
      "If you looking for an accommodation check our this website and if you want to rent a car in New York, I recommend this site.\n",
      " With over 220 designer stores like Dolce & Gabbana, Gap Outlet,…\n",
      "02\n",
      "American Dream\n",
      "Attention shoppers: the Greater New York City area is now home to one of the largest malls in the nation – the American Dream mall.…\n",
      "03\n",
      "The Hamptons, Sag Harbor & Outlet Shopping Day Trip from NYC\n",
      "A really great day trip is this trip to the famous Hamptons on Long Island.\n",
      "Query: top attractions in new york state\n",
      "Tavily Response: Discover the best things to see and do in New York State, from iconic landmarks like the Statue of Liberty and Niagara Falls to scenic parks and lakes. Explore the diverse regions, cultures, and history of this state with our guide.\n",
      "Tavily Response: Niagara Falls. #1 in Best Places to Visit in New York State in 2024. The tumbling, frothy falls - more than 3,100 tons of water flows over the edge every second - are the star tourist ...\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Draft: \n",
      "Based on the information provided, here is a suggested day trip itinerary for your group of 4 friends residing in NYC:\n",
      "\n",
      "Place: The Hamptons, Sag Harbor & Outlet Shopping\n",
      "Dates: Any weekday or weekend\n",
      "Estimated spending: $150-$250 USD per person\n",
      "Mode of commute from NYC: Private car or rental car\n",
      "Estimated time to reach destination: 2.5-3 hours\n",
      "\n",
      "Itinerary:\n",
      "\n",
      "Day 1: \n",
      "\n",
      "   - 8:00 AM - Depart NYC and head to the Hamptons\n",
      "   - 10:30 AM - Arrive in Sag Harbor, a charming historic whaling village. Explore the quaint downtown, browse the boutiques, and visit the Sag Harbor Whaling & Historical Museum.\n",
      "   - 12:30 PM - Have lunch at one of the local restaurants, such as The Page at 63 Main or The Beacon.\n",
      "   - 2:00 PM - Drive to the Tanger Outlets in Riverhead. Spend a few hours shopping at the various designer outlet stores.\n",
      "   - 5:00 PM - Head to one of the local wineries in the Hamptons, such as Wölffer Estate Vineyard, for wine tasting and enjoying the scenic vineyards.\n",
      "   - 7:00 PM - Depart the Hamptons and head back to NYC.\n",
      "   - 9:30 PM - Arrive back in NYC.\n",
      "\n",
      "This day trip will allow your group to experience the picturesque Hamptons, explore the quaint town of Sag Harbor, indulge in some outlet shopping, and enjoy the local wineries. The Hamptons offer a relaxing and scenic getaway from the bustling city, catering to a variety of interests within your group.\n",
      "\n",
      "Let me know if you would like me to modify or adjust the itinerary based on your group's preferences or any additional information you can provide.\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Critique: \n",
      "Thank you for the detailed itinerary for a day trip to the Hamptons from NYC. Overall, the plan seems well-structured and covers the key attractions in the area. However, I have a few suggestions to further improve the itinerary:\n",
      "\n",
      "1. Timing Adjustments:\n",
      "   - The driving time from NYC to Sag Harbor is likely closer to 2.5 hours, rather than 3 hours. This will give you more time to explore the town.\n",
      "   - Allocate more time (at least 3 hours) for shopping at the Tanger Outlets, as it is a large outlet mall with many stores to browse.\n",
      "   - Consider extending the winery visit to allow for a more leisurely experience, perhaps 1.5-2 hours.\n",
      "\n",
      "2. Dining Options:\n",
      "   - While The Page at 63 Main and The Beacon are good recommendations, you may also want to consider other lunch options that offer a more authentic Hamptons dining experience, such as the Sag Harbor Whaling & Historical Museum Café or the John Jermain Memorial Library Café.\n",
      "   - For the winery, see if you can find one that offers a food menu or light bites to accompany the wine tasting.\n",
      "\n",
      "3. Additional Activities:\n",
      "   - If your group is interested in outdoor activities, consider incorporating a visit to one of the Hamptons' beautiful beaches, such as Coopers Beach in Southampton or Main Beach in East Hampton.\n",
      "   - Depending on your group's interests, you could also explore other historic sites or art galleries in the Hamptons.\n",
      "\n",
      "4. Alternative Destination:\n",
      "   - While the Hamptons is a great option, you may also want to consider visiting the North Fork region of Long Island, which is known for its wineries, quaint towns, and a slightly more relaxed atmosphere compared to the Hamptons.\n",
      "\n",
      "Overall, the itinerary provided by the vacation planner is a solid starting point, but incorporating these suggestions could help enhance the day trip experience for your group. Let me know if you have any other questions or need further assistance.\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Queries and Response:\n",
      "Query: day trip from nyc to hamptons with updated itinerary\n",
      "Tavily Response: Hampton Bays makes a great day trip to the Hamptons from NYC with both LIRR and Jitney options for transportation if you don't have a car. Hamptons Bays makes for another great beach day from the city, grabbing fresh seafood from one of the many restaurants, or some fun watersports! You can rent bikes in the Hampton Bays, there are multiple ...\n",
      "Tavily Response: The first stop of the day trip to the Hamptons from NYC was Riverhead where we had a quick breakfast. From there the bus took us to the charming town of Southampton. We stopped at Coopers Beach for a quick 20-minute beach break before we headed towards Meadow Lane. This impressive street is known for its expensive houses and mansions of local ...\n",
      "Query: day trip from nyc to north fork long island\n",
      "Tavily Response: At Pugliese Vineyards, you can taste four wines of your choosing for $10. At Sparkling Pointe, the Grand Tasting Flight is four designated sparkling wines for $20. To get to the North Fork by train, take the LIRR from New York to Riverhead, Mattituck, Southold, or Greenport. It's a bit faster to drive.\n",
      "Tavily Response: The North Fork is a 30-mile-long peninsula in the northeast part of Suffolk County, New York, U.S., almost parallel with an even longer peninsula known as the South Fork, both on the East End of Long Island. The name comes from the fact that at Riverhead, Long Island, the peninsulas split into two like a fork (look at the map for the split!).\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Draft: \n",
      "Based on the information provided, here is a suggested day trip itinerary for the group of 4 friends residing in NYC:\n",
      "\n",
      "Place: The Hamptons, Sag Harbor, and Outlet Shopping\n",
      "Dates: One day trip\n",
      "Estimated spending: $200-300 USD per person\n",
      "Mode of commute from NYC: Private car or LIRR train\n",
      "\n",
      "Itinerary:\n",
      "\n",
      "Day 1: \n",
      "\n",
      "9:00 AM - Depart from NYC and drive/take the LIRR train to the Hamptons. The drive will take approximately 2-2.5 hours, while the train will take around 2.5-3 hours.\n",
      "\n",
      "11:00 AM - Arrive in Southampton. Start your day with a stroll through the charming town, browse the boutiques, and enjoy the scenic waterfront. \n",
      "\n",
      "12:30 PM - Head to Coopers Beach, one of the most famous beaches in the Hamptons. Spend an hour or two relaxing on the beach, swimming, or taking a leisurely walk.\n",
      "\n",
      "2:30 PM - Drive or take the bus to Sag Harbor, a historic whaling village with a quaint, picturesque Main Street. Explore the shops, art galleries, and grab lunch at one of the local restaurants.\n",
      "\n",
      "4:30 PM - Visit one or two of the North Fork wineries, such as Pugliese Vineyards or Sparkling Pointe, for a wine tasting experience.\n",
      "\n",
      "6:30 PM - Depart the Hamptons and head to the Riverhead Tanger Outlets. Spend a couple of hours browsing the various designer and brand-name stores.\n",
      "\n",
      "8:30 PM - Depart the outlets and head back to NYC, arriving around 10:30 PM.\n",
      "\n",
      "This itinerary allows your group to experience the scenic beaches, charming towns, and unique shopping experiences of the Hamptons and North Fork region, all within a comfortable day trip from NYC. Let me know if you would like me to modify or expand on any part of this suggested itinerary.\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Critique: \n",
      "Thank you for the detailed day trip itinerary to the Hamptons, Sag Harbor, and Riverhead Outlets. Here are my thoughts and suggestions:\n",
      "\n",
      "1. I generally agree with the overall structure and timeline of the itinerary. It covers the key highlights of the Hamptons region and provides a good balance of activities within the day.\n",
      "\n",
      "2. However, I would suggest reconsidering the order of the activities. It might be better to visit the wineries earlier in the day, before heading to the beach. This would allow you to enjoy the wine tastings when you're fresh, rather than at the end of the day.\n",
      "\n",
      "3. You may also want to consider adding an additional stop in Sag Harbor. The town has a rich history and a charming waterfront that could warrant more than just a brief exploration. Perhaps you could incorporate a lunch stop in Sag Harbor, instead of visiting during the afternoon.\n",
      "\n",
      "4. In terms of the Riverhead Outlets, while it's a convenient stop, you may want to consider allocating a bit more time, as 2 hours might feel a bit rushed for a shopping experience. Alternatively, you could suggest that the group returns to the outlets on a separate day, as it's quite close to NYC and can be easily accessed.\n",
      "\n",
      "5. Lastly, I would recommend providing some backup transportation options, in case the group decides not to drive themselves. Perhaps include information on shuttle services, Uber/Lyft availability, or even renting a minivan or van for the day.\n",
      "\n",
      "Overall, the itinerary is well-thought-out and covers the key highlights of the Hamptons region. With a few minor adjustments, it could be an excellent day trip for your group of friends. Let me know if you have any other questions!\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Queries and Response:\n",
      "Query: updated hamptons day trip itinerary with winery visits before beach\n",
      "Tavily Response: Hampton Bays makes a great day trip to the Hamptons from NYC with both LIRR and Jitney options for transportation if you don't have a car. Hamptons Bays makes for another great beach day from the city, grabbing fresh seafood from one of the many restaurants, or some fun watersports! You can rent bikes in the Hampton Bays, there are multiple ...\n",
      "Tavily Response: The first stop of the day trip to the Hamptons from NYC was Riverhead where we had a quick breakfast. From there the bus took us to the charming town of Southampton. We stopped at Coopers Beach for a quick 20-minute beach break before we headed towards Meadow Lane. This impressive street is known for its expensive houses and mansions of local ...\n",
      "Query: transportation options for hamptons day trip from nyc\n",
      "Tavily Response: Hampton Bays makes a great day trip to the Hamptons from NYC with both LIRR and Jitney options for transportation if you don't have a car. Hamptons Bays makes for another great beach day from the city, grabbing fresh seafood from one of the many restaurants, or some fun watersports! You can rent bikes in the Hampton Bays, there are multiple ...\n",
      "Tavily Response: During the peak summer seasons, trains frequently run around 6 times per day. However, during the off-season, they may only run once or twice a day at most. A round trip from NYC to Hamptons train on the LIRR costs $5.50. If you don't purchase a round trip, it will be $6.50 for the total cost since a $0.50 fee will be applied each way.\n",
      "Query: suggested additional stops in sag harbor\n",
      "Tavily Response: Sag Harbor, New York is a beloved destination for travelers seeking a quick Hamptons getaway. Read on for everything to know about a day trip to Sag Harbor.\n",
      "Tavily Response: The Hamptons most charming and laid back town might be Sag Harbor. This former whaling port has held on to its historic past and attracts more laid-back visitors. Read on for the top things to do in Sag Harbor, New York.\n",
      "**********************************************************\n",
      "**********************************************************\n",
      "Draft: \n",
      "Based on the information provided, here is a suggested day trip itinerary for the group of 4 friends residing in NYC:\n",
      "\n",
      "Place: The Hamptons, New York\n",
      "Dates: Saturday, June 10th\n",
      "Estimated spending: $150-$250 USD per person\n",
      "Mode of commute from NYC: Train (Long Island Rail Road)\n",
      "Estimated time to reach destination: 2 hours \n",
      "\n",
      "Itinerary:\n",
      "\n",
      "Day 1: Saturday, June 10th\n",
      "- 8:00 AM - Depart NYC via the Long Island Rail Road (LIRR) from Penn Station to Hamptons Station (Westhampton)\n",
      "- 10:00 AM - Arrive in Westhampton and take a short taxi/rideshare to Quogue Wildlife Refuge\n",
      "- 10:30 AM - Explore the Quogue Wildlife Refuge, a 300-acre nature preserve with hiking trails, birdwatching, and educational exhibits\n",
      "- 12:00 PM - Lunch at the quaint Westhampton Beach village, with options like The Lobster Grille or Sen Restaurant\n",
      "- 2:00 PM - Head to Sagaponack and visit Channing Daughters Winery for a wine tasting and tour of the vineyards\n",
      "- 4:00 PM - Spend time at Coopers Beach in Southampton, one of the Hamptons' most renowned beaches\n",
      "- 6:00 PM - Dinner at one of the Hamptons' farm-to-table restaurants, such as Stone Creek Inn or Calissa\n",
      "- 8:00 PM - Return to the Westhampton LIRR station and take the train back to NYC\n",
      "- 10:00 PM - Arrive back in NYC\n",
      "\n",
      "This itinerary allows your group to experience the natural beauty, charming villages, and local cuisine of the Hamptons region, with a mix of outdoor activities, wine tasting, and beach time. Please let me know if you would like me to modify or adjust the itinerary based on your group's preferences.\n",
      "**********************************************************\n"
     ]
    }
   ],
   "source": [
    "thread = {\"configurable\": {\"thread_id\": \"4\"}}\n",
    "output = []\n",
    "streamOptions = {\n",
    "    'task': \"Suggest me a day trip for a group of 4 friends all residing in NYC.\",\n",
    "    \"max_revisions\": 3,\n",
    "    \"revision_number\": 1,\n",
    "}\n",
    "for s in graph.stream(streamOptions, thread):\n",
    "    output.append(s)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### UI using GRADIO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "from helper import ewriter, writer_gui"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on local URL:  http://127.0.0.1:7860\n",
      "\n",
      "To create a public link, set `share=True` in `launch()`.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><iframe src=\"http://127.0.0.1:7860/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "MultiAgent = ewriter()\n",
    "app = writer_gui(MultiAgent.graph)\n",
    "app.launch()"
   ]
  },
  {
   "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": "3.12.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}