Epic 3.7 - Create the Curriculum Coach Assistant

[Ahmedr275]

Curriculum Coach Assistant Implementation

Overview

The Curriculum Coach is a specialized AI assistant designed to assist educators in curriculum planning and refinement. This assistant will provide tailored support for brainstorming, organizing, and improving educational content, reducing educators' workload and enhancing their teaching strategies. It leverages the existing AI chat framework and builds on the CoTeacher architecture to deliver focused functionality for curriculum development.

Purpose

The Curriculum Coach aims to:

• Assist educators in designing new courses or enhancing existing curricula.
• Automate repetitive tasks like resource curation and content summarization.
• Align courses and lessons with educational standards and objectives.
• Save educators time by providing actionable, AI-driven recommendations.

Features and Functionality

Core Functionalities

1. Suggested Prompts:

   • Examples:
     • "Help me outline a new high school biology course."
     • "What are key topics for a coding bootcamp?"
     • "How can I make my geography course more engaging?"

2. Standards Mapping:

   • Align content with educational standards for compliance.

3. Chat History:

   • Save and retrieve conversations for future reference and iteration.

Steps to Implement

Follow the Guide to Creating an AI Assistant on Marvel to create the Curriculum Coach assistant.

1. Folder and File Setup:

   • Create a folder: /app/assistants/curriculum_support/curriculum_coach.
   • Inside this folder, create the following:
     • core.py
     • assistant.py
     • /prompt folder for the assistant's system message.

2. Assistant Implementation:

   • Write a context-specific system prompt for the Curriculum Coach. The prompt should focus on curriculum planning, brainstorming, and resource curation.

3. Core Functionality:

   • Implement the executor function in core.py to:
     • Define the chat_context variable for contextualized prompts.
     • Handle high-level assistant functionality and return AI results.
   • Ensure core.py contains a user_info variable and a list of messages for each assistant.

4. Assistant Registration:

   • Register the assistant in /app/assistants/utils/assistants_config.json with:
     • Group: curriculum_support
     • Assistant Name: curriculum_coach
     • Path: curriculum_coach/core.py

5. Testing:

   • Test the assistant locally using a POST request to http://localhost:8000/assistant-chat with appropriate payloads.
   • Additionally, deploy the assistant using App Engine or Cloud Run. Share the deployment link and verify its functionality with collaborators.

Performance Benchmarks

To ensure the Curriculum Coach meets expectations:

• Evaluate its ability to generate accurate and actionable content for curriculum development.
• Assess the relevance of suggested resources and prompts.
• Measure response times for various actions.
• Collect feedback from educators for further improvements.

Acceptance Criteria

Functional

• The Curriculum Coach assistant is created within the /app/assistants/curriculum_support folder.
• Assistant-related files (assistant.py and core.py) are implemented and functional.
• Suggested Prompts and Standards Mapping are operational.
• Context-specific system prompt is accurately tailored for curriculum planning.

Integration

• The assistant is registered in /app/assistants/utils/assistants_config.json.
• All core functionalities integrate seamlessly with the backend.

Testing

• Unit tests for:
  • Context prompt generation.
  • Chat history integration.
• Manual tests for:
  • Resource curation and standards alignment.
  • Saving and retrieving chat history.

Additional Notes

• Use placeholder data for testing during development.
• Ensure the assistant can be extended for future improvements like advanced resource curation and analytics.

How to Test

1. Deploy the assistant locally and send POST requests to http://localhost:8000/assistant-chat.
2. Validate the assistant's responses for course creation, resource curation, and standards mapping.
3. Confirm chat history is saved and retrieved correctly.
4. Collaborators must deploy the assistant using App Engine or Cloud Run, provide the deployment link, and test the functionality.

Resources

• Designs
• Guide to Create an AI Assistant on Marvel
• Assistants README