20260629 - AI Enabled Delivery for Consultants

Overview

This Areopa webinar (61 minutes, published 2026-06-30) shows how a functional consultant can use AI across the project lifecycle. Demos use Microsoft 365 Copilot to turn workshop notes into a requirements worksheet with Business Central interpretations and open questions, and to generate UAT test cases with coverage mapping, negative cases and multiple user roles.

The session also covers critiquing solution designs, standardizing documentation with custom instructions and saved memories, and issue summarization and root cause analysis. It then introduces Business Central’s built-in Copilot features, MCP, and Copilot Studio agents. The speaker stresses that the consultant stays responsible for validating output. The Q&A covers Copilot cost and AI code generation.

Key points

• Copilot in Excel turned workshop notes into a requirements worksheet with Business Central interpretations, source mappings and an open questions section; the consultant remains responsible for accuracy.
• Prompts work best when they define a persona, context, scope and instructions; Copilot can also critique a proposed solution, suggest alternatives and question whether custom development is really required.
• Copilot generated UAT test cases with preconditions, test data, steps, expected results, variations and a coverage map to requirements. About 30 workshop points took roughly a minute; the speaker rates them a first draft that gets you 40 to 50 percent of the way in minutes rather than hours.
• Custom instructions and saved memories in Microsoft 365 Copilot personalization settings can enforce a consistent documentation structure and style, and can be toggled on or off per task.
• MCP lets AI agents work with Business Central APIs; an administrator chooses which entities to expose. Consultants may need to define which data an agent can see and do, whether it is read-only or action taking, and who approves risky outputs.
• Microsoft 365 Copilot costs about $30 per user per month, varying by country; most of what was shown runs in it.
• On code, the speaker relays developers reporting a real uplift since the start of 2026; the moderator noted that vanilla or well-componentized code gives better results, while an attendee reported a poor experience with complex customizations.

Features

AL objects mentioned

As heard in the captions. A name that matches one object page by exact type and name links to it; the others stay as named.

• other “Microsoft 365 Copilot” at 9:33
• other “Power Apps” at 10:22
• other “Microsoft Learn” at 11:07
• other “worksheet” at 13:19
• other “Excel workbook” at 24:15
• other “UAT test cases worksheet” at 25:36
• other “Copilot Studio” at 57:37

Quotes

• 4:26 “some very practical techniques you can apply immediately to enhance your productivity as a functional consultant through the use of AI”
• 5:29 “show me how this AI stuff can actually help me survive real world workshop specs testing and indeed go live pressure”
• 8:04 “accelerated requirements gathering is is an area where AI can definitely assist you”
• 8:50 “A lot of it is just working through very well identified and known stuff. But really, if you use AI tooling, you should be”
• 13:19 “Microsoft 365 copilot can see obviously external sources but it can also see anything within my business that I have access to in terms”

Disclaimers in the video

• 4:51 other: This is not going to be a deeply technical session. It’s not going to be about cutting edge possibilities and …
• 14:13 other: You’re still fully responsible for making sure that this is accurate and that you action these properly
• 40:41 coming-later: future agentic scenarios future it could be very near future where AI helps execute repeatable business tasks
• 52:31 other: not aware of any any indications that that’s going to change regarding copilot pricing

Presenters (as heard): Luke, Nile McGovern, Eager.