The AI Business Analyst is here — and it’s not replacing you, it’s working with you.

In this episode, Benjamen breaks down the Top 10 ways AI, and especially Microsoft Copilot, is already transforming the BA role. From requirements elicitation and process modelling to risk management and benefits tracking, we’ll look at how secure, enterprise AI tools can become your daily co-pilot.

This is about the AIBA — the AI Business Analyst. The next step in our evolution as BAs.

Episode Transcript

(00:00):
Now we need to talk about something that's not just
theory, it's already here. It's in my day-to-day work.
It's going to reshape how we alloperate as business analysts.
I'm calling it AIBA, the AI Business Analyst, not because AI

(00:21):
will replace us just now, but because AI is becoming our
copilot today. And the very best example of
this action right now is Microsoft Copilot.
When it's enabled inside your organization, it's locked down

(00:41):
and secure. You can feed in your own
documents, requirement packs, design decks, strategic papers,
and Copilot can analyze them, summarize them, or even rewrite
them. Not perfect.
It's a bit of clunky, you often have to go 1 document or image

(01:01):
at a time, but it works and if you learn to use it, you're
already stepping into the futureof business analysis.
The Better Business Analysis Institute presence, the Better
Business Analysis Podcast with Kingsman Walsh.

(01:25):
Welcome back to the Better Business Analysis podcast with
your host Benjamin Walsh. And we're getting into my top
ten ways AI and specifically tools like Copilot, Microsoft
Copilot, A transforming the BA role.
Now we've talked about AI and how to use it as ABA.

(01:45):
This is literally a role that's going to develop.
There will be people advertisingfor a IBAS in the future or
whatever they call it, so I wantto talk about it from that
angle. Number one is requirements
elicitation. With the AIBA, Copilot can turn

(02:06):
raw stakeholder input into structured requirements.
You can upload your meeting transcripts or workshop notes.
I took photos of my notes while talking to someone this week,
and I asked Copilot to extract the key points from that and

(02:27):
give me some rough user stories.OK or epic.
I wasn't going to just use thoseverbatim, but it did a lot of
the steps for me. And after a stake at a workshop,
Copilot can produce a draft requirements pack in minutes if
you want. I mean, that's the sell.
It's where we want to get to. It's not quite there.

(02:48):
Deloitte uses AI driven transcriptions and Copilot
summaration during their client workshops and they generate
requirements documents for review straight away.
Now we're not quite there and Deloitte has fed in the
templates they want to use the output they want.
This week I asked Copilot to summarize some diagrams I had

(03:10):
and put it into a pack for me using the branded template of
the client. And I gave it the branded
template of the client. It knew about it, but what it
did is it just gave me some really random template.
So it's not perfect #2 improvingdocumentation quality. copilot

(03:32):
can rewrite your BA output to beconsistent and professional
language and help with spelling or other issues that you might
have. I mean, I'm not the best with
grammar and copilot understands that.
There's no more messy drafts andyou know, version two and three.

(03:54):
And you can get style alignment within BA standards if you fed
it all that. You've got to be careful how you
prompt. I took some scrappy draft I did.
I asked copilot to rewrite this in context with some other
content I'd already written, andit lined it up, you know?
Again, not perfect, but saved meat least half an hour.

(04:18):
Microsoft's own product teams use Co pilots internally to
publish technical requirements documentation and they do that
into like executive ready language.
So that's angle you could come on at and #3 is the data
analysis at scale. This is so important.
So Copilot can query your internal data sources directly

(04:42):
if you're allowed, if it's configured, if it's secure, and
it can work with Excel Power, BISQL and soon Microsoft Fabric.
So there's your whole pipeline and you can ask natural language
questions. I've used copilot with Fabric,
Microsoft Fabric, which is the latest tool in data analytics
for pipeline management. And we can ask questions around

(05:04):
student numbers. It's very, very, very useful.
You could ask which product lines mis forecast by more than
10% in quarter 2 without having to do the data analysis.
It literally will go and find that answer for you.
Heineken uses Microsoft Copilot across Excel and Power BI to
analyse sales figures from thosetwo sources.

(05:27):
So you know the Excel's bridge hasn't been put in the data
warehouse yet. And it gives managers insights
that usually take days #4 is theprocess modeling and
optimization. I'm not that happy with the
output I'm getting from AI in this area.
But copilot can analyse your process documents and suggest

(05:50):
improvements. You can upload your existing
process maps, Visio maps, Imagesof Visio maps work better.
You can ask copilot where are the bottlenecks or duplicates.
It will do that. You can feed in like a customer
service SOP and it will highlight redundant steps.

(06:10):
So it does the kind of lean 6 Sigma on top of it.
Siemens have partnered with a couple of companies including
Microsoft to automatically map and optimize supply and chain
processes from the ERP logs thatthey've got.
Now I'll just on that point, yeah, I'm not happy with the
process diagrams. If you want to top tip, ask

(06:32):
Copilot to generate you what's called a mermaid diagram, which
is a syntax, basic syntax and itwill and you can take that and
then that will be your very simple diagram.
It's not BPMN compliant. So we're not the maybe one of
you can work with me on making atool like that.
Number 5 is stakeholder communications support.

(06:55):
So copilot can reframe BI deliverables for different
audiences. You have one draft and you have
the exec version, the developer detail client facing.
So you do start with your biggest pack and then you can
ask Copilot to rewrite those points and you just have to copy
and paste them into another presentation.
And it does take a little while to still design those documents,

(07:19):
but it saves you hours just rewriting and you can say
summarize this 20 page design pack into a one page executive
level brief and it will do that.PwC does that, uses their, their
consultants use it to tailor client communications on the fly
and they turn that around and they also present back to board

(07:42):
ready reports, again using templates, which they'll
probably work in Microsoft on number six is test cases and
acceptance criteria. It's really good at this.
So Copilot can generate test cases straight from the
requirements if you've done a good job with them.
It can draft scenarios. It supports QA alignment with BA

(08:06):
intent. So it it like helps match those
two, which is always a gap. And you can just upload your
requirements doc and say generate acceptance criteria in
each cases for a tester and it will do that.
JP Morgan are piloting using generative AI to basically come

(08:28):
up with the test automation scripts from business
requirements. So that's pretty amazing #7 is
that benefits realization tracking?
It's actually really good at that.
So you can copilot can connect BA requirements to
organizational KPIs, right? And I do that via the capability
map, which I've used this week. You need to ingest, you know,

(08:51):
project financials and status reports and you can track
benefits and see if they're materializing.
You could say based on these project reports, costs reducing
at the target. Are we on track over in the UK,
the NHS, which is a National Health Service, the free doctors

(09:13):
and hospitals over there, they use Google AI analytics to
measure whether or not digital health projects are succeeding
or not #8 is that Copilot can surface hidden risks within your
requirements, packs and plans. It can flag vague requirements.

(09:34):
It can challenge you like as a, as a, as a friend would or
someone who's reviewing your document work.
And it can highlight dependency clashes that you may not have
picked up, right. And you can ask, what are my
ambiguities in the requirements document?
Where can you identify some risks?
Where do I need to add more detail?

(09:55):
Airbus, you know, the plane manufacturer uses AI modelling
to assess risks in supply chain projects.
And that spots things that humans may have missed #9 is BA
training and knowledge sharing. Copilot becomes kind of an
internal BA mentor we just mentioned.

(10:18):
So new hires can actually query past project packs, which is a
good one. Define consistency.
Things went well. Lesson learned.
Copilot can surface patterns andthose lessons we talked about,
you could say, show me past CRM project requirements and their
outcomes for this organization. What's worked well, Accentia has

(10:41):
actually developed AI coaches for their new consultants,
right? Instead of having a people lead
or a capability lead. And they provide context
specific guidance from internal knowledge bases.
And actually this week Microsofthas released a new AI tool that
sits across SharePoint for your records management, which is
around knowledge management. So check that out and #10 the

(11:04):
future of BA is to work with AI every day.
Copilot can be your secure partner.
And there are other tools. I've just used Microsoft
Copilot. They work generally in a
Microsoft environment. And Microsoft's very good at
targeting, you know, my kind of role.

(11:24):
AI handles the grunt work. Beos focus on facilitation,
translation and value and turn around quickly, much quicker
than we do today. And so in the future, we will be
working very closely with AI andAI may do some of our tasks, but
we really need to be Aibas first.

(11:46):
And I was sure you'll start seeing some jobs out there that
are focused specifically on this.
OK, so there's my top ten. We're entering the era of the AI
business analyst or that AIBA, they're not replacing us yet,
but sitting alongside us. Tools like Microsoft Copilot are

(12:10):
already doing this today securely inside organizations
with your own documents. Sure, it's not always slick,
sometimes it's one document at atime, but the point is it works
and it's only going to get better very quickly, like in a
couple of weeks. So if agile was the length from
Waterfall, then the IABA is the next leap and you'll start

(12:34):
seeing these roles. So give it a try.
Use AI as your copilot. Build your muscle in promoting,
validating and partnering with these tools.
That's how you future proof yourself as a better BA.