Project Description

Inspiration 🌟

The inspiration behind Iris stemmed from the need to revolutionize how small businesses manage client interactions and automate order processing. Many small businesses face challenges in keeping up with customer inquiries and processing orders efficiently, which negatively impacts the customer experience. Iris addresses this gap by leveraging advanced AI technology to create an intelligent chatbot that simplifies communication, streamlines order creation, and enhances operational efficiency for businesses.


What It Does 💬📦

Iris is an AI-powered agent designed to optimize small business operations. It acts as a virtual assistant that:

  • Communicates Seamlessly: Provides real-time responses to customers using AI agents for efficient interaction.
  • Automates Order Creation: Handles order workflows by automatically generating and managing them in Directus CMS, simplifying inventory and order tracking for businesses.
  • AI-Driven Engagement: Employs SambaNova's Meta-Llama-3.1-70B-Instruct model to interpret customer needs, offer precise responses, and autonomously manage order processing.

How We Built It 🛠️

To bring Iris to life, we utilized:

  • Directus CMS: As a centralized system for managing products and orders with an intuitive interface for businesses.
  • SambaNova Cloud: Powered by Meta-Llama-3.1-70B-Instruct, delivering ultra-fast, real-time AI processing.

  • Al Agent Workflow:

Al Agent Workflow

  • Streamlined Order Workflow: Seamless integration of SambaNova's AI with Directus for efficient and automated order creation, ensuring accuracy and speed.

Challenges We Faced 🚧

Developing Iris presented a few hurdles:

  • API Integration: Ensuring reliable communication between the AI agent and Directus CMS for accurate data management.
  • Performance Optimization: Fine-tuning the AI's response times to ensure real-time interactions without lag.
  • Complex Workflows: Designing a robust, automated process to handle diverse customer needs while maintaining simplicity for businesses.

Our Achievements 🎉

We’re proud to have created a transformative solution for small businesses:

  • 🕒 Time Efficiency: Iris automates client interactions and order creation, significantly saving time for business owners.
  • 🤝 Enhanced Engagement: Offers a consistent and professional experience for customers, increasing satisfaction and trust.
  • 📈 Streamlined Processes: Successfully automated the journey from inquiry to order completion, integrated with Directus CMS.

What We Learned 📚

Through this project, we gained invaluable insights into:

  • AI-Driven Integration: Using SambaNova's advanced AI models to power intelligent workflows and manage complex operations.
  • Building Efficient Systems: Developing autonomous processes that cater to both user and business needs.
  • Real-Time Interactions: Optimizing AI responses for seamless, human-like communication.

What’s Next for Iris 🚀

Our future roadmap includes:

  • 🌐 Enhanced AI Personalization: Implementing tailored product recommendations to boost customer satisfaction.
  • 💬 Multi-Platform Integration: Expanding Iris’s reach by incorporating it into various communication channels.
  • 📊 Advanced Business Insights: Adding analytics tools to help businesses understand customer behavior, trends, and preferences.

Hackathon Alignment

Participation in the Lightning Fast AI Hackathon

For the Lightning Fast AI Hackathon, we focused on demonstrating how high-speed AI can transform small business operations.

  • Challenge Addressed: Current AI systems struggle with real-time processing. Using SambaNova’s Meta-Llama-3.1-70B-Instruct, we created an AI agent capable of lightning-fast responses, optimizing order management for small businesses.
  • What Sets Iris Apart:
    • Fully automated order workflows within Directus CMS.
    • Real-time responses powered by SambaNova’s cutting-edge AI model.
    • A tangible application showcasing the potential of rapid, scalable AI agents for real-world use cases.

How to run IRIS locally

Clone the repository:

git clone https://github.com/yesid-lopez/iris-agent

Remember to set the env variables in the .env file.

Build the Docker image:

docker build -t iris:1.0 .

Run the app:

docker run -p 8501:8501 iris:1.0

Built With

  • cms
  • directus
  • langchain
  • langgraph
  • llama
  • python
  • sambanova
  • streamlit
Share this project:

Updates