This FAQ covers how to use Kadal Content Manager and Kadal AI for content development, agent creation, and content analysis. Find answers to common questions about using Kadal effectively, creating and managing agents, and following best practices.
Q1: What can I do with Kadal Content Manager and Kadal AI?
A: With Kadal, you can:
- Upload Content on Kadal Content Manager: Use Kadal Content Manager to store and manage documents, multimedia, and more.
- Upload Canvas Courses on Kadal AI: Import courses from Canvas for AI-driven content analysis.
- Create Agents in Kadal AI: Develop agents that assist with content development and analysis.
- Call Agents in Kadal Content Manager: Apply agents directly in Kadal Content Manager to analyze or enhance content.
- Develop and Analyze Content: Use Kadal Content Manager in tandem with AI agents to create, improve, and assess content.
Q2: How do I create effective agents in Kadal AI?
A: Here are some best practices for creating agents:
- Focus on Micro-Tasks: Agents work best when assigned a single, specific task in content development or analysis.
- Keep Instructions Simple: Avoid overwhelming the agent with too many guidelines or tasks.
- Test Thoroughly: Test agents repeatedly in Kadal AI to ensure they perform accurately.
- Use Concise Prompts: Short, clear instructions in the agent’s knowledge base and prompts yield better results.
- Request Deletion If Necessary: If an agent isn’t meeting expectations, request deletion through the helpdesk.
Q3: What are the main content development use cases for Kadal AI?
A: Kadal AI supports a range of content creation tasks, including:
- Creating Course Objectives: Develop clear, measurable objectives in a consistent format.
- Designing Course Structure: Outline a course structure that aligns with objectives.
- Generating Lesson Content: Produce full-text content for units, modules, or lessons.
- Developing Multimedia Elements: Create assets like storyboards, flashcards, and slideshows.
- Creating Assessments: Design multiple-choice questions (MCQs), scenario-based questions, and project-based assessments.
- Formatting Content: Structure and organize content for better readability.
Q4: What are the main content analysis use cases for Kadal AI?
A: Common uses for content analysis in Kadal AI include:
- Analyzing Course Objectives: Review objectives for clarity and alignment with standards.
- Checking Content Quality: Use AI bots like Readability Checker and Style Guide Bot to maintain quality.
- Evaluating Course Quality: Select course sections and review based on quality benchmarks.
- Analyzing DEI Compliance: Assess content sections for Diversity, Equity, and Inclusion (DEI) standards.
Q5: What should I keep in mind while creating agents?
A: Follow these Do’s and Don’ts:
Do’s
- Understand the Content: Familiarize yourself with the source material, guidelines, and expected output before creating an agent.
- Learn from Existing Agents: Review successful agents to understand their structure and prompts.
- Ensure Human Oversight: Always assign someone to review and verify AI outputs.
- Get Permissions for Proprietary Content: Seek PM and client permission before uploading proprietary materials.
- Contact Helpdesk for Assistance: Reach out if you encounter issues or need help with agent configuration.
Don’ts
- Avoid Unnecessary Details: Only include essential instructions in the agent prompt.
- Don’t Reuse Prompts Blindly: Tailor every prompt for new agents.
- Don’t Expect Perfection Instantly: AI agents may need fine-tuning to achieve desired results.
- Don’t Upload Protected IP: Avoid uploading intellectual property without permission.
- Don’t Assume Issues Are Permanent: Report any system issues to the helpdesk for support.
Q6: How should agents be tested in Kadal?
A: Agents should be continuously tested until they meet performance expectations. Only publish an agent if you’re fully satisfied with its accuracy. If needed, seek helpdesk assistance to delete unsatisfactory agents.
Q7: What if I want to delete a published agent?
A: To delete a published agent, raise a request with the helpdesk, as only they can facilitate this action.
Q8: How can I ensure quality and compliance in AI-generated content?
A: Implement quality checks by:
- Testing agents regularly and adjusting prompts as needed.
- Assigning a human reviewer to evaluate AI output.
- Using AI tools like readability and style checkers for additional quality assurance.