Portfolio Assignment 3

Note: M3 - Group Assignment 3 Deadline: Monday, 24th Feb at 12:00 PM

Introduction

This assignment is designed to enhance your understanding of modern AI techniques, particularly the use of AI agents in conjunction with GPT models. You will learn to build and deploy AI workflows using FlowiseAI and CrewAI, creating practical solutions that demonstrate the power of these technologies. Your project will culminate in an interactive Streamlit app to showcase your results.

Task Description

Create an innovative project that employs GPT models and AI agents to solve a real-world problem of your choice. The guidelines are intentionally broad to encourage creativity and exploration. Here are some project ideas to inspire you:

1. Research Assistant Agent

  • Build an agent that helps researchers gather and analyze information from multiple sources
  • Implement document summarization and key insight extraction
  • Create an interactive Q&A system about the collected information
  • Visualize relationships between different pieces of information

2. Content Creation Suite

  • Develop a system of agents that collaborate on content creation
  • One agent for research, one for writing, one for editing
  • Include tools for tone analysis and style consistency
  • Add features for generating accompanying visuals or presentations

3. Personal Productivity Assistant

  • Create an agent that helps with task management and organization
  • Implement smart scheduling and priority assessment
  • Add features for breaking down complex projects into manageable steps
  • Include progress tracking and reporting capabilities

4. Data Analysis Companion

  • Design an agent that helps users understand and analyze datasets
  • Implement automated data cleaning and preprocessing
  • Add features for generating insights and visualizations
  • Create natural language interfaces for data querying

5. Learning and Education Agent

  • Build a personalized learning assistant that adapts to user needs
  • Create interactive lesson plans and study materials
  • Implement knowledge testing and progress tracking
  • Add features for generating practice exercises and explanations

Data

  • You may utilize datasets from 🤗 Hugging Face, Kaggle, or create your own.

Delivery

  1. GitHub Repository

    • Create a repository containing your code, notebooks, and documentation.
    • Include a README.md with a brief description of your assignment, your approach, and instructions on how to run your code.

  2. Colab or Jupyter Notebook

    • Save your working notebook in the repository.
    • Ensure that the notebook includes detailed explanations of each step, from data preprocessing to model evaluation.

  3. Group Work

    • You may work in groups of up to 3 members.
    • Each group member’s contribution should be briefly outlined in the README or the notebook.

  4. Technical Explainer Video

    • Record a short (~5 minutes) technical explainer video presenting your main ideas, methodology, and results.
    • You may use Panopto, OBS Studio, Loom, or any other screen-recording tool.
    • Include the video link in your submission.

  5. Submission

    • Send an email to Hamid (hamidb@business.aau.dk) with the link to your GitHub repository (and video) by the deadline.

Good luck, and let your creativity shine! 🎉