Effective AI Prompting Techniques in D365 Finance

1. Introduction
2. ROLE — “Who is the AI?”
3. TASK — “What should be done?”
4. CONTEXT — “What should the AI know?”
5. OBJECTIVES — “What does success look like?”
6. STOP CONDITIONS — “When is it done?”
7. OUTPUT FORMAT — “How should it be structured?”
8. FULL PROMPT EXAMPLE (D365)
9. ADDITIONAL PRACTICAL USE CASES
   1. 1. Finance (D365 Finance)
   2. 2. Supply Chain Planning
   3. 3. Data & Reporting
10. Conclusion

Introduction

A good AI response depends less on the AI itself and more on how you ask.

Think of prompting like giving instructions to a new team member: the clearer and more structured you are, the better the result.

This guide breaks prompting into simple steps, with concrete examples including Dynamics 365 Finance & Supply Chain Management (D365 F&SCM).

ROLE — “Who is the AI?”

Why it matters: The role defines the level of expertise and perspective.

Bad example: Explain inventory management.

Good example: Act as a senior Dynamics 365 Finance & Supply Chain consultant specializing in inventory optimization and warehouse management.

D365 Example: Act as a D365 Supply Chain expert with experience in implementing advanced warehouse management (WMS) for manufacturing companies.

TASK — “What should be done?”

Why it matters: A vague task leads to vague answers.

Bad example: Help me with D365.

Good example: Design a warehouse process in D365 to optimize picking and reduce errors.

D365 Example: Design a step-by-step process in D365 SCM to implement batch tracking and traceability for a food manufacturing company.

CONTEXT — “What should the AI know?”

Why it matters: Context transforms generic answers into relevant solutions.

D365 Example:

Company: mid-size manufacturing company 

Modules used: Inventory + Warehouse Management 

Constraint: must comply with traceability regulations 

Pain point: frequent picking errors and stock discrepancies 

OBJECTIVES — “What does success look like?”

Why it matters: Defines priorities and decision criteria.

D365 Example:

Objective:

– Reduce picking errors by 30%

– Improve stock accuracy

– Ensure compliance with traceability regulations

STOP CONDITIONS — “When is it done?”

Why it matters: Prevents incomplete or overly long answers.

Example:

Stop when:

– A complete process is defined

– Each step is linked to a D365 feature

– Risks and benefits are explained

OUTPUT FORMAT — “How should it be structured?”

Why it matters: Well-structured outputs are easier to use.

Example:

Provide:

– Step-by-step process

– Table with: Step / D365 Feature / Benefit / Risk

– Final recommendations

FULL PROMPT EXAMPLE (D365)

Act as a senior D365 Finance & Supply Chain consultant.

Task:

Design an inventory management process to improve stock accuracy.

Context:

– Manufacturing company

– Using D365 SCM

– Issues with stock discrepancies

– Limited team training

Objectives:

– Improve stock accuracy

– Reduce manual corrections

– Ensure scalability

Stop conditions:

– Full process described

– Each step mapped to D365 functionality

– Clear recommendations included

Output format:

– Step-by-step process

– Summary table

– Final recommendations

ADDITIONAL PRACTICAL USE CASES

1. Finance (D365 Finance)

Prompt:

Act as a D365 Finance expert.

Design a process to automate vendor invoice posting using OCR and workflows.

2. Supply Chain Planning

Prompt:

Act as a supply chain planner using D365.

Create a demand forecasting approach using historical data.

3. Data & Reporting

Prompt:

Act as a Power BI expert connected to D365.

Design a dashboard to track inventory turnover and stock aging.

Conclusion

A strong prompt is:

• Clear
• Structured
• Contextualized

With practice, prompting becomes a powerful skill to:

• Save time
• Improve decision-making
• Unlock real business value from AI