How To Use AI To Improve Your Clothing Brand In 2025 — Harish Chandramowli | Why Data Integrity Mistakes Cost Fashion Brands Millions, Why Most Fashion Business ERP Implementations Fail, How AI Visualizes Complex Processes For Clothing Brands (#419)

In this episode, Harish Chandramowli, Co-founder and CTO of Flaire, explores how AI is transforming the fashion and clothing industry—especially in streamlining operations and improving data management. 

He highlights the challenges brands face when relying on spreadsheets, and stresses the importance of data integrity for better decision-making. 

He also emphasizes the need for flexible pricing and the importance of adapting technology to fit each brand’s unique workflows in the fashion and clothing industry. 

Topics discussed in this episode:  

• Why spreadsheets are secretly killing fashion brands' profitability. 
• How AI is revolutionizing fashion operations. 
• What data integrity mistakes cost fashion brands millions. 
• Why most fashion ERP implementations fail. 
• How to sanitize messy fashion data. 
• What technical flexibility really means. 
• Why pricing strategy determines software adoption. 
• How AI visualizes complex fashion processes. 
• What workflow understanding reveals. 
• Why collaboration beats technology. 

Links & Resources 

Website: https://www.Flairesoftware.com/
LinkedIn: https://www.linkedin.com/in/scharish/
LinkedIn: https://www.linkedin.com/company/Flaire-software

Get access to more free resources by visiting the show notes at
https://tinyurl.com/3c2b4vj4

MORE RESOURCES

Subscribe to our FREE Newsletter: https://newsletter.ecommercecoffeebreak.com/

Free Store Optimization Beginners Guide: Instant PDF Download! 👉 https://ecommercecoffeebreak.com/essential-conversion-optimization-guide/

Rate & Review: Help others discover the show by rating the show on Apple Podcasts at https://podcasts.apple.com/us/podcast/ecommerce-coffee-break-digital-marketing-podcast-for/id1567749422

Advertise on ECB - https://ecommercecoffeebreak.com/podcast-sponsorship/

Follow the podcast to catch all the bonus episodes I am adding. Hit that follow button now!

Transcript

Hello and welcome to another episode of the Ecommerce Coffee Break podcast. Running a fashion brand is tough. If your operations still live in spreadsheets, things can go off track very fast. Missed orders, stuck issues, and way too much manual work will hold your work back. So it's very, very difficult. So we want to dive into this topic today. And with me on the show, I have Harish Chandramouli.

And we want to talk about and look into how AI can help you to clean up the chaos and run a smoother and smarter business. Harish is the co-founder and CTO of Flaire, an AI-powered ERP platform built for fashion brands. Flaire helps turn messy operations into clean and scalable growth. And we want to dive into this right now. So let's welcome him to the show. Hi, Harish, how are you today?

Harish Chandramowli (03:22.306)
Good.

Claus Lauter (03:24.219)
Harish, what does a typical day look for a fashion brand stuck in spreadsheets?

Harish Chandramowli (03:30.315)
So some of the fashion brands who we moved from spreadsheet to flare, some of the highlights they say are like one.

When there are too many spreadsheets, they miss coordinating between them because the data gets replicated in different spreadsheet. An example for a typical example would be a purchase order manager knows when you are going to receive an item, what date clothing is going to come. And then there will be a marketing person who would be coordinating like a marketing campaign. that means you, before you list online, you need a model to come in. You need some photo shoots, you need to book the dates. And that needs to be like the next day of

of the sample arrives because you don't want to give the difference. That means lost space and sale. And when there are multiple spreadsheet, one person updates the PO manager updates the spreadsheet where you track the productions and like income in inventories, but then the marketing campaign manager just doesn't see it.

But when everything is in one system, for example, in these cases, one person enters the data, everyone gets different views similar to spreadsheet, but the data just flows through everywhere. That means you don't miss the dates. You don't miss the updates. So those are some of the things I have heard from fashion brands on how they operate. But also, again, everyone's day-to-day work within fashion brands is very different. When you are small, there are two or three people who does everything, which actually is easier to coordinate. Everyone has everything in their mind.

And as you grow you have a CEO who makes business decision who sees how profitable we are not just like when the items are coming in and then makes decisions around there and then you have a planner who needs to understand how much items are coming in when they are coming in as well as like how much we are selling or we are certain item faster. Also imagine in fashion you have like a very specific time period a quick time period to sell because once the season is out a season is like three to four months

Claus Lauter (04:57.531)
Thank

Harish Chandramowli (05:26.704)
Once the season is out, you are sitting on the inventory. Those clothes are not going to move out. So the planning becomes more and more critical and important in fashion industry. So you have a planner who does that. And then your procurement manager or production manager who kind of negotiates with factory, especially with tariff, all the numbers changing. This takes a higher precedence these days. How do you figure out where to get the right inventories? And then you have like marketing, PR, like different people. But yes, it's mostly like.

someone starts wearing a jacket like Jack of all trades and as the company expands you bring in more people.

Claus Lauter (06:04.687)
Now, as you said, small brands might start as a one-man show and then the spreadsheet is fine because you have everything in the back of your mind. But as you grow, things become more complicated. How did your past work with data? And you're coming from that side. Help you understand these problems.

Harish Chandramowli (06:20.93)
So one, I'm not a fashion person first of all.

me being in fashion industry was something I never imagined five, six years back, I guess. I worked at Bloomberg, MongoDB, some of the very famous data platform and helped build some of the core products. And I was basically in the store called OANS and like Soho and heading to their back office meetings, kind of looking at like talking to their founder and just observing how they are operating. And I realized that back office and operations of fashion is fundamentally a data and

workflow problem.

And one, don't want, just because it's the problem I like, and I've been doing it in different industries for the past decade, I don't want to jump in. So I was looking at the existing solutions. One of the things I realized was like Shopify has done an amazing job with e-commerce and the front end of fashion, storefronts and everything. And then if you are like a hundred million plus brand, NetSuite is like a gold standard. But there are a lot of SMB brands who can't pay for NetSuite, who can't pay for the customizations needed for NetSuite. And they were still sitting

spreadsheets and they were missing these updates there often it's like they are not getting the most out of the data and coming from data background that's what made me the most curious like how can I bring in innovation for this SMB market without making them pay the 100k customization fees that an H-Suite charges. I built

Harish Chandramowli (07:49.078)
If for people who do not know MongoDB is like a database where it holds all the data, I build some parts of like their serverless database. And usually in database context, we see everything as X and numbers. But when I was talking to fashion industry, you have like colors, UPC codes, as like this kind of text and numbers kind of primitive data tapes. That's what people use the terms people use every day. And color is nothing but like a bunch of like X code, which is a characters and then like name and everything.

How can I extend this to make it flexible and bring innovation so that a database itself can hold these kind of fashion-specific data? And that led us to be like, if we have like a...

analyze like a female athletic wear brand and like a TA3 like a female swimwear brand when they use the platform they have like cup size torso length regular sizes and then color to form a skew whereas we have like men's fashion wear brand O &S which I mentioned about like men's fashion wear who just sees their product as like t-shirt is just color and size that's it so we made it flexible that based on your business how you want to see your business how you want to analyze your business

you can just fill an Excel sheet of this is the properties I need, boom, the software will come up. So that helps you make the business decision faster. A lot of things I explained are technical and how innovation came in, but at the end of the day, as a user, you get a product where you view it very similar to spreadsheet. You get the columns that you use right now to make business decisions. That's pretty much the crux of.

how to bring SMBs the data that they can use. They can get the best out of those data.

Claus Lauter (09:36.155)
It makes perfect sense. you said, fashion industry, there's just different data points that you don't have in other industries and you're stuck in a system that might not really serve your purpose in running your own business there. tell me data. Obviously nowadays data equals AI. Because that's the fastest way to crush the data to get output that you can really use. How does your system help on that front?

Harish Chandramowli (10:05.592)
So I can talk about AI in two ways when it comes to ERP. One easier part everyone relates is that, I have this data. I need to ask questions around what is selling fast, what is not selling fast. What are the things that might mean replenishment? Remember, one of the things is that AI cannot just go and keep ordering in fashion industry because of the seasonality I talked about. January of this month might be colder than January of last month. So people's purchasing.

methods differ. That means it's not like a typical other retail industry where you could just ask like I want to compare my last year's sales data and order it right because that is like unknown factor that comes into play. But what people want to do is like hey last year at these temperature weather situations in these months what did I sell? Which product I sell? Then you can ask the AI like now I want to compare similar products and then make planning decision. technically where I

coming at us like maybe it's not that to take over a planners whole workflow but it is going to enable planners to be more efficient. That's an easier part of AI. feel like so many companies are trying to do it and

If you are in Shopify, there are not just Flaire, right? There are so many other people out there. The other part of it, which I observed throughout the process, is that for these people to make decisions, for AI to make decisions, you need a...

data that's trustworthy. And that's where everything lives in spreadsheet and people don't enter the data because softwares are not easy enough for them to use. And that's where I also feel the bigger disruption in AI can happen to help you make better decisions. Where like you can build custom workflows. can change your Shopify with like extra parameters, extra properties that you want to track. With AI building those customizations are very easy. So I think even ERP as a whole,

Harish Chandramowli (12:03.148)
have to go and be like this is the exact functionality and I need to use the ERP rather you could now go to come to me and say like this is my workflow I need a product that solves it so that I enter the data from day one from the planning phase till when I get the data to when it's sold that way the other part of the AI which uses these data

can easily work because that has so much innovation. It's so easy to do. So that's why we are concentrating more at the beginning stages because every one of our customers, we want to make sure they have the data. And then, yeah, you can make those decisions. So yeah, at the end of here, I want to appeal to your office is more like.

world is changing. So to a certain extent, you could go to people and be like, I don't care about technology. I don't care about software. This is my workflow. Give me something that works for my workflow that your company needs to change for using a software.

Claus Lauter (12:58.811)
like the approach and I think a lot of listeners will be happy to hear that because everyone is building a business and every business has a different workflow, a different style of how they manage the business, the ordering process and so on and so forth. And if you have a system that basically adapts to what you do and not just a ERP system that you have to adapt to, then that's a huge advantage, a huge benefit there. Now, tell me, from your experience with your clients, how's the process from coming

from a cluttered spreadsheet world into going into an ERP system? What kind of data or what kind of homework did they do to get started?

Harish Chandramowli (13:39.438)
So for me, it is two different types of companies. Alala, for example, are using a very outdated ERP for over 10 years. And when we want to bring them on, they only add data.

after a factory confirms it that they are shipping. The only reason they had the data from that point in ERP is because you need a key in every value and it's just pain when something changes. It's not hooked up to your email system to just send POs out of the platform. That's with legacy ERP. So when we moved over, the challenge was making sure that whatever is in the work in progress data also moves into the system from their spreadsheet. That moment was actually, I would say,

like very seamless because there was an ERP, there was a certain structure there. It's easy to move them to a new platform. And I have seen that when people are already using an ERP, it's easy to move them.

When people that are not using an ERP when they are only in spreadsheet, they do not have any kind of structure, especially at the beginning stages, for us to put them in a way where we can run reports. Yes, we can put them one on one from spreadsheet, and you can't run reports. And sometimes, even barcode assignments, for example, when it's not tracked through a system, a barcode was assigned to three different.

skews basically like barcode is the UPC code that you get and it needs to be unique for each item whereas they were like replicating so the data sanitization part definitely took a lot more time than i would have liked but definitely all these custom AI tooling to help you with those but to certain extent i can't do anything right like a human has to say that okay now that i have assigned the same barcodes for three different items which one do i choose

Harish Chandramowli (15:29.164)
When I was looking at these people, when I was onboarding these people, the first thing I got is that like, don't even have to wait for...

need to be 10 people, 20 people to board like any full-fledged ERP system. Even at the beginning stages, even with Shopify, don't have a free flow text. You can choose from a drop down. So you need to take some time to see how to configure your Shopify so that there is data integrity. Data integrity is like the biggest piece, even for AI to be more useful. So one of the advice I have is like if you are a new fashion brand that's, or any retail brand that's starting out, and if you are still in spreadsheet,

That's OK. At least for a one-on-one use case, make sure lot of these color, namings, and everything has a convention. lot of them are like even within your spreadsheet is a dropdown rather than each user typing it in. That was the biggest challenge I have seen when I moved people from spreadsheet to an ERP. This is like someone who is already using an ERP for 10, 20 years.

Claus Lauter (16:29.957)
And it makes perfect sense. I think that gives also the merchants a chance to revisit the data structure they have. And as I said, you need to have data that is really quality data to make it work. And I think with a process going from looking into the data and what you're doing now into a system like yours, there's a lot of advantages there in just getting better in the quality of data that you work with. Now.

Harish Chandramowli (16:52.748)
Yes, so one of the exercises I do with all my merchants, which I feel is so much easier with AI, is that I ask them like, hey, explain to me all the process you do. What are the different roles in the company and what is everyone's like day to day and what they see in their spreadsheets. Then feed this to AI and it will immediately draw like a nice flow chart. Previously it used to take me two hours to put in effort to do flowchart. This like immediately. Then you walk through which is like, this is your flowchart. This is how your business is.

operating, this is what you need at each stage. Now let's take a step back and see how to structure your data so that each and every one can be very efficient in their day-to-day tasks. And that homework is really important. Even if someone comes and sells you a dream ERP, I feel if you don't do the homework that you're just going to pay the ERP and not get the best out of it.

Claus Lauter (17:49.775)
Yeah, and think it also helps to have someone coming from the outside with a fresh set of eyes and helping you to look at your business because it's easy to get lost in your own business and don't see the forest because of the trees. So having somebody not only providing the software like you do, but also providing a fresh view on what you're doing helps a lot. Can you share some examples of brands that you have worked with and what kind of results they saw?

Harish Chandramowli (18:15.884)
Here we swim for example, cut out bunch of...

They used to have someone to upload backing list, which is the things that you get from factory. So we have an AI that now they can just upload, which passes and they just need to verify. So you don't need another up-work person doing it. Another brand used to have an up-work person to create labels, just like these size labels, make it in a nice size and send it to factories. Now they don't have it because the labels just get generated immediately. And another PO manager, this is the most feedback I've got from every customer.

is that their PO managers are more efficient now. Because we have like very simple functionality of like send to vendor that immediately like changes your whole PO to PDF, sends it through in the backend through your own Gmail. That is nothing changes in their part. That has made them like more than 50 % efficient. It used to take them like one hour to write a single PO, download it, upload in your Gmail, the, write the email and send it to your factory. Now it happens in seconds, not like the two hour of

whole process. And even more importantly one of the biggest thing I heard is not just the time savings. We have a CRM part where you could upload a document and then use the, even HubSpot doesn't do it which becomes a fashion thing like you can create like a outreach with an attachment. Even in HubSpot you can't have an attachment and that is one of the

biggest complaint I heard from people is because now a lot of times when they do these bulk emails, when you have to attach while sending the email, people attach the wrong flyers. Like, hey, my employee made a mistake by attaching the spring 2024 flyer instead of a fall 2024 flyer. So now what happens is that data integrated come into play. The CEO of the small brand uploads the flyer, creates a template that this is going to be the outreach for the next three months. So users just click their customers and send the email.

Harish Chandramowli (20:13.292)
is not just making people more efficient, it's making sure your employees do not make any mistakes. So there are two possibilities. Everyone is efficient, the biggest feedback I always get is like, I don't see mistakes from any of my employees. I can trust the data I see in this platform. I can trust that my workflow just will work.

Claus Lauter (20:30.555)
you

Claus Lauter (20:35.695)
Perfect. I want to dive a little bit on the technical side of things. Obviously, you said data connection is very important, data flow is important. What kind of interfaces and API does Flaire offer?

Harish Chandramowli (20:47.118)
So we

Very on technical side is like we follow something called BF of pattern because my backend is very different from front end. So we are by default like even within the UI, you see it operates out of an API. So if you have APIs, we would immediately like kind of open up our APIs endpoint that you can build any integrations you need. From the front end side, this allows me to be more nimble. So based on users workflow, I can create like custom UI, custom workflow, other user interface changes for each and every customers.

And it's easy. From a technical perspective, why this makes sense to me is because you want frontend to go with the flow of every customer's workflow and not like people have to adopt. But when it comes to testing, when it comes to stability of the platform, your back end needs to be very robust. So it cannot be like you write back end code for each customer.

So what we do is like, whether it's a PO, data management and everything, it's all our backend opens up the APIs and we make sure all those are built, tested out a manual QA just apart from all these automations and testing. So I can trust that my backend works very well. Frontend.

What I observed is like all my customers, once they say this is the clicks I do, they do not do anything different. They do the exact same clicks. So it's easy to build a custom front end and be like, OK, these clicks are working. My customer is going to be happy.

Claus Lauter (22:13.049)
OK, now that sounds very good. mean, having a stable platform in the back end is very important, but also having a platform that's user friendly is important. Now, we were already touching a little bit on the onboarding process. But can you tell me what kind of steps are involved and how long does it take to get up and running?

Harish Chandramowli (22:29.046)
It varies. Again, if you're coming from an existing ERP, it's usually like within one to two weeks, I can move you because your data is in like a very good state. But when you are not coming from an existing ERP, I tell people, let's not rush through this before you sanitize your data. Then you would regret it later. So let's take like whatever time you need based on how your data is to cleanse it. Because cleansing the data needs to come from the customers. So they need to make sure they share it. The process, how it usually works is like I first

talk to them like I said I draw the flowchart to understand their business and be like let's both are in the same let's both of us be in the same page one I want you to be sure that this is how your business operates two I'm an outsider to fashion industry the way I learn about each customer is by talking to you about is this how your business works so it's learning on both sides right

And once that is done, sometimes I have to explain to them to restructure their architecture. An example would be for Alala, we did this and it was great. So, traditionally ERP, someone places an order in website. It goes to your ERP and from ERP it goes to your warehouse. And sometimes the connection breaks. used to take them 12 to 24 hours sometimes to get the order to warehouse. But you don't need that. The truth is Shopify to warehouse integrations are so nice.

Why don't we integrate warehouse and Shopify? But Flaire, like traditional ERP, give you the data by observing what happens. So when a customer places the order, five minutes, your warehouse people have it. Within an hour, you ship it. But you get the data in Flaire. So I kind of need to educate them a bit on like, I know this is how your traditional ERP has taught you that everything has to happen through ERP. And it slows you down.

As someone who comes from a very technical background, I see data and flow very differently. Why can't we re-architect and make ERP less critical? But operationally, it's like you are more efficient.

Harish Chandramowli (24:32.203)
So that education sometimes I feel like takes anywhere from a week to a month based on the customers, based on how much time I need to explain and like set up. For some people, I don't need to set up a demo. I explain this, they get it. For some people, I have to set up like a demo environment to say why this is more efficient in long run, why this is a good thing to have. Even if you cut out FLAr tomorrow, you are still operational and you can try different ERPs.

And yeah, it kind of takes sometimes, like for some people to set up the demo and explain to them what a need takes like a month.

Claus Lauter (25:09.787)
I totally can relate to that. So the fashion brands that I have worked with, the usual founder is not a data scientist. That's what I can tell you. So they're coming from a different mindset. And I think it's good to have somebody like yours on the side who might ask even uncomfortable questions to get your business optimized and you get the data flow right. again, a fashion brand, they're not necessarily coming from crunching numbers on AI level or ERP level. So that's a huge benefit there.

Now, how does your pricing structure work?

Harish Chandramowli (25:41.742)
Like most ERPs, very honestly, the pricing is pretty flexible based on the business and the amount of module you need and data. So we work through it. One, I strongly believe pricing should not be based on users.

because then that defeats the purpose of everyone being in the system and using the system every day, entering the data every day. So we don't do usage-based pricing. The way we structure the module is pretty much like how much SKUs you have and whether you just write POs or whether you need CRM modules. So the base price is, I would say $10,000, but our customers have anywhere, sweet spot has been like anywhere from $20,000 to $50,000 based on your operations.

Claus Lauter (26:28.347)
That obviously gives you security in your budgeting and you know what you're dealing with. So that's perfect.

Harish Chandramowli (26:33.132)
Yeah, we try to be very transparent. We don't charge for implementation. And I also mean, it's a startup. At this point, I tell my customers, if we end up seeing that it's going to be more than what I promised you, I wouldn't do anything for the first or second year. But at the third year, we can again negotiate on what is the right thing to do.

Claus Lauter (26:52.891)
Okay, Aarhus, before our coffee break comes to an end today, is there anything you want to share with our listeners that we haven't covered yet?

Harish Chandramowli (27:00.302)
If you're a fashion business hearing this and if you are in Shopify, do not have ERP. Even if you don't want an ERP, reach out to me in LinkedIn. I'm happy to just geek out on how you can use Shopify effectively from day one. Because that way when you want an ERP, you are in a good place. That has always been the takeaway for all the fashion brands I kind of like speak with.

Claus Lauter (27:24.251)
Okay, cool. Where can people go and find out more about you?

Harish Chandramowli (27:27.246)
I am very active in LinkedIn. You can look up SC, Harish, Santa, Canada, Henry, Apple, Rad, guess, Indiana, Santa and Henry.

Claus Lauter (27:41.403)
Cool. I will put the links in the show notes anyway, so you will be just one click away. And I hope a lot of our listeners will reach out to you. think it's an amazing combination that you have there, coming from a data perspective, working only with fashion brands. I don't think there's anyone else on the market doing this. There's a lot of advantages and benefits in there, in working with you. Thanks so much for your time today. I hope to hear from you soon again. And all the best.

Harish Chandramowli (28:09.816)
Thank you.