When using AI systems, what are some best practices for ensuring the results you receive are accurate, relevant, and aligned with your original goals?

1. Be Specific: Instead of a broad request, provide context (who you are), the task (what you need), and constraints (length or tone).
2. Give Examples: Show the AI 1–2 examples of the style or format you want to see.
3. Ask for Reasoning: Use the phrase "think step-by-step" to help the AI process complex logic and reduce errors.
4. Refine and Iterate: Treat the first response as a draft; give the AI feedback to tweak the tone or fix details until it's perfect.

Garbage in garbage out - good GAI output needs detailed and specific input.

When working with AI systems, following a few structured best practices can significantly improve the accuracy, relevance, and usefulness of the results. Here are key practices that help ensure AI outputs stay aligned with your goals.

Define exactly what you want the AI to accomplish before writing the prompt.

Example

• Weak prompt: “Analyze this project.”
• Strong prompt: “Identify the top three risks in this project plan and suggest mitigation strategies.”

Clear goals reduce ambiguity and guide the AI toward the right output.

AI performs best when it understands the background, constraints, and audience.

Include details such as:

• Industry or domain
• Target audience
• Project scope or timeframe
• Desired tone or format

Example:

“Summarize this project update for executive leadership in three bullet points focusing on risks, milestones, and next steps.”

Large or vague prompts can produce inconsistent results. Instead, divide the task into smaller prompts.

Example workflow:

1. Ask for a summary
2. Ask for risk analysis
3. Ask for recommendations

This improves clarity and control over the output.

AI responses should be treated as assistance, not final truth.

Best practices:

• Verify facts using reliable sources
• Compare with internal data or documentation
• Review outputs for logical consistency

This is especially important for technical, financial, or strategic decisions.

Prompt engineering is iterative. Adjust prompts based on the output.

You can refine by adding:

• More constraints
• Desired format (table, bullets, summary)
• Role-based instructions

Example:

“Act as a project manager and analyze the following sprint backlog for delays.”

Requesting explanations can help you evaluate whether the answer is logical.

Example prompt:

“Explain how you arrived at this recommendation and list any assumptions made.”

Tell the AI exactly how you want the response structured.

Examples:

• Bullet points
• Table
• Step-by-step plan
• Executive summary

This ensures results are easy to interpret and usable.

Avoid sharing confidential or sensitive information unless the system is approved for secure use.

✅ In summary:

To get the best results from AI systems:

• Define clear goals
• Provide context
• Break tasks into steps
• Validate outputs
• Iterate prompts
• Request structured formats
• Apply human judgment

These practices turn AI from a simple tool into a reliable decision-support system.

It is important to check if the data which has been provided is correct then measure the outcome, does it include all the aspects covering cost, schedule and quality, timelines, qualitative & quantitative risks etc.

Garbage in, garbage out.

Cite the sources and follow CREATE while generating request.

Using AI is a challenge when not using the prompt accurately, just like you are not able to present your thoughts in front of a meeting.