3 Ways to Use AI in Excel: Chatbot, Sidebar or Template.
Dashboards, analytics
and templates in Excel, designed with AI.

The cover is a short version of the blog
title. Sample dashboard, fictional data.
You do not need
to run Excel yourself. Let AI help you design it, and let Excel do the work.
That is the idea of this post, and the same idea runs through my Tamil AI book.
|
Books by the author AI
for the Rest of Us
- a practical guide to using AI at work (English). Ebooks on Amazon | Google Play Books ஃபோட்ரான்
இன்ஜினியரும்... ஏஐ சாம்பாரும்! - a simple AI book in Tamil. Ebooks on Amazon | Google Play Books For hard copies, contact me through the blog. |
Read this
article in Tamil: எக்செல்லை நீங்கள் இயக்க வேண்டாம், ஏஐ இயக்கும் உங்களுக்காக கச்சிதமாக
9 min
read
In this post
1. Is it magic by AI, or by
Excel?
3. Different ways of using AI
in Excel
4. A Live Example: One
Template, Not 10 Files
Short of time? Read Section 4 first, then come back for the rest.
1. Is it magic by AI, or by Excel?
We use Excel
every day, in families, offices and shops. In the banks and finance offices I
have known, SQL servers and big systems exist, yet many people still do much of
their work in Excel.
Microsoft
expanded the number of rows, from 65,536 to 1,048,576 in Excel 2007, most
likely to meet popular demand. A worksheet can now hold 1,048,576 rows. To me, that shows how much
users depend on Excel. People simply keep adding rows instead of moving to
Access or SQL. This is what I have seen in my career.
I have used
Excel from Lotus 1-2-3 to MS 365. As MIS head at Sutherland Global, I handled
about 400 people at one point. In my team, nine in ten spent most of their
working day in Excel, cutting, pasting and scrubbing data, day in and day out,
and making a clear dashboard for the CXO team.
Ready-made
tools like Power BI and Crystal Reports exist, and they are good. But you still
have to configure what you want, and make it look right. That takes time.
For a smaller
team there is another way. Describe the report in plain words to an AI, and get
a working template with the tiles, charts and summary lines already in place.
Is it magic
by AI, or by Excel?
2. The idea
In database
design there is a simple rule: enter a figure once, and let everything else
read from it. This is called normalisation: each fact lives in one place
only. Our Excel files rarely follow it. We start with two columns, add two more
later, and keep adding rows and columns at will. Soon the same number sits in
three places, and nobody trusts the total. Garbage in, garbage out.
This is where
AI gives relief.
AI can help you
design the file properly before you start. It sets up one place for data entry,
one place for calculation, and one place for the dashboard. After that, Excel
does the calculating with ordinary formulas. AI suggests the layout, the
drop-down and the charts. Excel gives the answer. AI designs, Excel
calculates.
Here is the part that surprised me. Take one sentence on my sample
dashboard:
“Oct
2026: revenue ₹17.91 lakh, down 5.9% on last month.”
To write that sentence correctly, the formula must ask two questions. Is
there an earlier month to compare with? If yes, did revenue go up or down? That
is two IFs, and three possible sentences:
|
Earlier
month available? |
Revenue
change |
Sentence
ends with |
|
No |
n/a |
“no earlier
month to compare with.” |
|
Yes |
0 or more |
“up x% on
last month.” |
|
Yes |
below 0 |
“down x% on
last month.” |
In my
experience, many people write one IF, test the case in front of them, and stop.
Then the first month of data shows an error, or a nonsense number. AI worked
out every case, the way a truth table does. In the whole sample, it wrote about
50 IF conditions in 45 cells. Many are the same formula copied down a column.
AI wrote
them. Checking them is our job.
3. Different ways of using AI in Excel
Each way has its merits and demerits. None is “the best”. Use the one
that fits the job.

Figure 1: Three ways to use AI with
Excel. Same data, three workflows.
The choice
is not which tool is smarter. It is where the numbers live.
In the first
two ways, you get an answer. In the third way, you get a file that keeps giving
answers. That is why it suits monthly work. You check the assumptions once,
with two or three sample entries, and the same file works next month without
starting again.
For the sidebar
way, test the version you have today. Tools change fast. What was a static
chart last year may be editable now. For example, Google announced in June 2025
that Gemini’s charts in Sheets are now fully editable,
where earlier ones were static images.
For live data
connections, very large data, and shared company-wide reports, Power BI is
still stronger. I use both for different jobs. For data already in Excel, this
third way is easier to start.
So the rule is
simple. One-time question, use Way 1. Work inside the same file, use Way 2.
Same work every month, use Way 3.
4. A Live Example: One Template, Not 10 Files
Every office
knows this Monday morning scene.
Manager asks,
“Where is last month’s report?” You open your folder and find Final.xlsx,
Final_v2.xlsx, Final_FINAL.xlsx - and still nothing is final.
This is where I
tried a different way.
This is a
sample template, built with Claude AI. You can make a dashboard like this for
your own job with a simple prompt. Treat this template as a starting point, not
a finished answer. You kick-start from here, then make it yours.
|
Download the sample template: Monthly Sales MIS Dashboard (Excel) Please open it in
desktop Excel. All data in it is fictional, so you can see how it works
before trying your own. |

Figure 2: The sample dashboard, with
fictional data. Pick a month and a target; the rest updates by itself.
What it does
Your original
data stays in one place - the data sheet. One-time entry only: date, region,
product, salesperson, units, revenue, cost.
AI designed the
interactive dashboard without a pivot table, customised for your need. You pick
a month from a drop-down and type a target. That is all.
The dashboard
then shows six Key Performance Indicator (KPI) tiles, six plain sentences, a
12-month trend against target, and three ranked charts. You get the full
picture in minutes, without opening 10 files.
The 10-versions problem
Many of us save
one sheet for each month, and end up with 10 versions. One template and one
drop-down replace all 10.
Let me say this
with my own work, fairly.
In my Mutual
Fund Distributor (MFD) client work, I use a Power BI sheet interlinked with
Excel files. The client wants to see the same asset composition in different
ways - star rating wise, fund house wise, category wise. In total, that becomes
10 sheets for 10 views.
Power BI is not
wrong here. It does the job. But the work is heavy. Excel files feed Power BI,
Power BI feeds the views, and any small change means checking both sides.
Now I suppose I
can do the same 10 views with one Excel file built with an AI template.
Same source
data. One drop-down. Pick star rating, or fund house, or category - and the
dashboard changes. No 10 files. No interlink to maintain.
I suppose it
will be simpler and faster.
I am not saying
Power BI should be thrown away. I am saying, for this one monthly job - seeing
the same data in 10 ways - a single Excel template may be enough. And if it is
enough, why carry 10 sheets?
That is the
practical question I want to leave with you: How many versions are you
maintaining today for one job?
A small technique worth knowing
The summary lines are not typed. They are formulas. Three things make it
work:
•
& joins text and
numbers into one sentence.
•
TEXT(number,"0.00")
formats the number inside the sentence. Without it, Excel shows 17.912 instead
of 17.91.
•
IF(change>=0,"up
","down ") switches the words depending on the figure.
For the “top
region” line, the sheet ranks the names first, then picks the first. A tiny
tie-breaker stops two equal values from returning the same name.
AI designs.
Excel calculates. You decide.
What to check before you rely on it
|
Please try it before you trust
it: •
Change
the month and the target, and watch everything update. •
Paste
a small month of your own data and compare one total with a figure you
already know. •
Keep
dates as real dates, not text. The sheet shows a red warning if it finds text
dates, text amounts, a name missing from the lists, or more than 20,000 rows. •
Read
how the AI built it. A formula can be wrong in an unexpected place. |
Where it may not fit
This is meant
for small and medium teams doing monthly reporting. This is about knowing where
to use the right tool.
5. Everyday use
The same method
works for monthly reports, project trackers, expense sheets, and to-do lists.
It works for mutual fund data from CAMS and KFintech turned into charts.
Students can
track study plans and job applications. A home can keep its monthly grocery and
school-fee accounts. A small business can track orders and collections.
If your need
changes, upload the template to AI and say what to change. That is all.
I have done
this for my own work.
Usually we keep
separate sheets, without any links. One sheet for book sales, one for
Print-on-Demand (POD) printing, one for marketing posts. Each sheet lives
alone. If one changes, the other does not know.
I sat with AI
tools and built it step by step. With my book-sales tracker, built with Meta
AI, it is different. It is live, and I use it regularly. Several sheets in one
template, all linked, work like one coherent engine.
A potential
order in one sheet becomes an actual order in the next, then delivered, then
paid - and every sheet knows the change.
So far, so
good. It is still my data, my logic. Only now, the sheets talk to each other.
The first
project always takes more time. Like the first sambar in a new kitchen - you
have to taste, adjust salt, and learn where the vessels are. The second time,
you already know the kitchen.
That is how I
see it. This first tracker taught me the method. Now the same method can be
used for day-to-day work in Excel.
I hope this
will grow into a bigger, simple transformation in how we use AI in Excel in the
coming days - not by replacing Excel, but by making Excel work for us.
6. Closing
Excel has been
our calculator for a long time. AI does not replace Excel. It helps us design
Excel better.
That is the
whole idea of this blog. AI designs, Excel calculates, you decide.
Download the
template here, try it with a few sample rows, and
change it for your own work. Do not start with your full live data. Start
small, check one total you already know, and then make it yours.
If it saves you
one hour next month, it has done its job. If it saves you 10 versions, it has
done more.
Tell me in
the comments - what is the one report you are still doing with 10 files?
Download this article as PDF
Good for
offline reading and sharing: PDF link
Read this
article in Tamil: எக்செல்லை நீங்கள் இயக்க வேண்டாம், ஏஐ இயக்கும் உங்களுக்காக கச்சிதமாக
We can work together
I can help you in two practical ways:
•
Making new templates using AI - for your monthly
reports, trackers, expense sheets, or mutual fund reporting. One template, one
drop-down, no 10 versions.
•
Learning to use AI in Excel - for your team,
your students, or yourself. Simple, tested method. No jargon.
Also: AI
Training & Awareness • Personal Mutual Fund Guidance • Soft Skills &
Career Guidance • Project Guidance for Students
For unbiased,
quality advice, reach me through Blog Contact Us. No spam, no pressure.
Connect with Kannan M
Blog: https://radhaconsultancy.blogspot.com/ | LinkedIn
| Instagram | Facebook | YouTube
| X / Twitter | Amazon
Author
Want to learn
AI in Excel for your own work? Send a message through Blog Contact Us.
About
the books
- read one, or read both. They are not duplicates.
The English and Tamil
books are not the same book translated. Each has its own examples, stories and
everyday use cases.
AI for the Rest of Us (English): Amazon | Google Play Books
ஃபோட்ரான் இன்ஜினியரும்... ஏஐ சாம்பாரும்! (Tamil): Amazon |
Google Play Books
For hard copies, contact
me through the blog.
Disclaimer: These are my
own ideas, polished with AI help for language. The template is a tested sample.
Please test it with your own data before relying on it for any money-related
decision.
#AIinExcel
#Excel #AI #ExcelDashboard #ExcelTemplates #MonthlyReporting #MISReporting
#AIForTheRestOfUs
No comments:
Post a Comment