saadtoorx/multi-agent-event-planner
๐๏ธ Multi-Agent Event Planner
An AI-powered Multi-Agent Event Planner built with CrewAI and Streamlit. This project demonstrates how specialized autonomous agents collaborate to research venues, coordinate logistics, and design marketing & communications for events โ producing a usable event plan and downloadable reports.
<div align="center"> <img src="images/demo_screenshot.png" width="80%" alt="Event Planner Demo Screenshot"> </div>
๐ Features
- ๐ค Multi-Agent Collaboration โ Distinct agents handle venue sourcing, logistics coordination, and marketing/communications so tasks are parallelized and specialized.
- ๐ End-to-End Event Workflow โ From venue research to marketing strategy and final downloadable summary.
- ๐ Web Research & Scraping โ Integrates search and scraping tools to gather venue and vendor details.
- ๐ง CrewAI Memory & Context โ Agents share context (task outputs) so downstream tasks use validated inputs (e.g., chosen venue feeds logistics & marketing).
- ๐งฉ Simulated or Real Execution โ Supports a simulated planning mode out-of-the-box and an option to execute with real CrewAI agents when configured.
- ๐ฆ Downloadable Artifacts โ Venue JSON, marketing report (Markdown), and a complete event summary are exportable.
๐๏ธ Architecture Overview
๐ง Agents
- Venue Coordinator โ Identifies candidate venues, checks capacity/availability, and outputs structured venue details.
- Logistic Manager โ Plans catering, equipment, transportation, and on-site logistics using venue constraints.
- Marketing & Communications Agent โ Designs promotional messaging, suggested channels, and a marketing timeline.
๐ Tasks
- Venue Task โ Research and select a venue that meets event requirements and produce
venue_details.json. - Logistic Task โ Coordinate catering, rentals, transport and create confirmations/checklist.
- Marketing Task โ Produce a marketing strategy and promotional materials as
marketing_report.md.
๐ ๏ธ Tools Used
WebsiteSearchTool/ SerperDev-style tool โ for web search and quick facts.ScrapeWebsiteToolโ extract structured details from vendor/venue pages.- CrewAI framework โ orchestrates agents & tasks.
- OpenAI API (LLM) โ agent reasoning, planning, and content generation.
- Fallback/mock tools are used when live search/scrape packages are not available so the app still runs.
๐ Project Structure
multi-agent-event-planner/ โโโ .github/ # ๐ GitHub Actions workflows โ โโโ workflows/ โ โโโ pushtohf.yml # ๐ Auto-deploy to Hugging Face Spaces โโโ app.py # ๐ Streamlit app interface (main) โโโ agents.py # ๐ค Agent definitions (Venue, Logistics, Marketing) โโโ tasks.py # ๐ Task definitions (venuetask, logistictask, marketingtask) โโโ tools.py # ๐ ๏ธ Tool configuration & fallbacks for scraping/search โโโ apputils.py # ๐ Helpers: API setup, printing utilities, etc. โโโ requirements.txt # ๐ฆ Python dependencies โโโ LICENSE # ๏ฟฝ Project license (MIT) โโโ README.md # ๏ฟฝ This documentation
๐ ๏ธ Installation & Setup
- Clone the repository
git clone https://github.com/saadtoorx/multi-agent-event-planner.git cd multi-agent-event-planner
- Install dependencies
pip install -r requirements.txt
- Configure API keys (required for real-agent execution)
Create a .env file with your keys:
OPENAIAPIKEY=your-openai-key SERPERAPIKEY=your-serper-key # Optional: set model name used in the app OPENAIMODELNAME=gpt-4o-mini
Note: If you don't provide search/scrape keys or CrewAI packages are missing, the app falls back to simulated execution so you can test the UI and outputs.
โถ๏ธ Running the App
streamlit run app.py
- Open the sidebar and enter OpenAI and Serper API keys to enable full functionality.
- Fill event details (topic, city, date, participants, budget, venue type) and click Start Planning.
- Choose Simulated or Use Real CrewAI Agents before executing the planning run.
- Download generated
venue_details.json,marketing_report.md, or the complete summary.
๐ฆ Tech Stack
- Python 3.8+
- Streamlit (UI)
- CrewAI (multi-agent orchestration)
- crewai-tools / SerperDev-like search & scrape tools
- OpenAI API (LLM)
- pydantic (output models)
- python-dotenv
๐ค Agent Summary
๐ฎ Future Improvements
- ๐ Integrate live vendor & venue APIs for booking checks and pricing.
- ๐งพ Add automated budget optimization and cost breakdown.
- ๐ Export plans to calendar formats (ICS) and PDF reports.
- ๐ Add UI visualizations for workflow, timelines, and resource allocation.
- ๐งฉ Add more specialized agents (e.g., Sponsorship Agent, Budget Analyst).
๐ License
Licensed under the MIT License. See LICENSE for full terms.
โจ Built By
Created with โค๏ธ by Saad Toor (@saadtoorx) Powered by CrewAI, Streamlit, and OpenAI