The New Creative Process

AI is transforming the creative process, helping designers move from ideas to concepts faster than ever. In design and retail, it is opening new possibilities for creativity, personalisation, and innovation. In this Entrepreneur UK interview, Sreeraman Mohan Girija, founder of Fynd, a retail technology company, talks about how AI is changing retail, from design and production to the way customers discover products, and why human judgement still matters. 

What problem did Fynd set out to solve?
We started in 2012 as Shopsense, with a fairly simple question: could technology make retail work better for the customer and the retailer at the same time? Back then a store in Mumbai could only serve the person standing at its door. Someone searching in Delhi or Bengaluru couldn’t see any of it. The inventory existed. The demand existed. What was missing was the connection between them. So the early work focused on discovery and engagement inside the store: helping a customer find what was actually available, and helping the retailer convert the footfall it already had. Solving that surfaced the bigger problem. Inventory, stores, orders and later the digital channels were all running in systems that didn’t talk to each other. Retail wasn’t short of technology. It was fragmented. Everything we’ve built since has been about closing that gap. We kept moving up the chain: inventory integration for Reliance Brands, then endless aisle, then marketplace fulfilment, then unified commerce. The real problem we’ve been solving for fourteen years is fragmentation. Every step of a retailer’s journey had a different vendor, and none of the data talked. Fynd exists so a brand can get design, catalogue, inventory, storefront, payments and delivery under one umbrella without replatforming what it already runs.

Why is the UK the right market for Fynd at this stage of its global expansion?
Because the UK is already past the point where digital is a side bet. More than 60%of retail sales happen online, one of the highest shares in Europe, so when something in your operations is slow, it’s not an IT issue, it’s a revenue issue. That’s the door we come in through. There’s also a very specific gap here. 80% of UK shoppers are already using AI tools to shop, and 68% can’t name a single AI shopping experience worth remembering. Everyone’s shown up and nobody’s impressed. That tells us there is still room to get the experience right. The technology in itself is not the constraint, but whether retailers have the right infrastructure to make it work. Only a fraction of retailers have the setup or the expertise to deploy AI end-to-end. I walked more than 100 stores in London recently and a lot of what I saw felt a decade old. We don’t think that should come down to deep pockets or a big in-house data team, so we come in as an AI-native layer on top of what a retailer already runs — fragmented, ad hoc systems to integrated, real-time operations, without replatforming. We’ve run this playbook at scale before, in India, the Gulf, Africa, Southeast Asia. For us, the UK is a natural next step and, scale what we’ve already built and proven in other markets. 

What are UK retailers asking for now that they weren’t asking for two years ago?

Two years ago it was “what should we pilot.” Now it’s “what does this do to my numbers.” That shift happened fast. But the more interesting change is what part of the business they’re asking about. It used to be the front end – the storefront, the assistant, the recommendation engine. Now they’re asking about the plumbing. Is my product data good enough? Is my stock accurate right now? Can an agent actually trust what my catalogue says? We had a deployment on a marketplace app with over 100 million downloads where someone searched for a black T-shirt and got a purple one back, because the word “black” was sitting in the wrong field. The catalogue beats whatever logic you train the agent on, every time. Nobody was asking us about that two years ago. It’s the first thing on the table now.

Fynd Create connects trend intelligence, design, sourcing and production. How does that change the creative process?

Think about how it works today. Trends in one tool, design in another, tech pack goes over the wall to sourcing, and three weeks later somebody comes back and says the fabric doesn’t work at that price. Start again. Every handoff loses something. What Fynd Create does is put all of it in one place. The platform is reading social signals, runway, competitor moves, the brand’s own data, and a designer can generate it off a prompt like “resort wear” while already seeing which of our 800-odd vendors and mills can actually make it, at what cost, by when. Fynd Snap handles the imagery so you’re not waiting on a photoshoot to fill the catalogue. The creativity doesn’t change. What changes is that the constraints show up at the beginning instead of the end. 

Your early deployments show up to 60% greater design productivity. What is driving that improvement?
Very little of it is AI making pictures. Most of it is dead time we took out. Concept work that used to take a week now takes an afternoon, so a team can look at ten directions instead of two. Tech packs and catalogue assets come off the design instead of being rebuilt by hand. Sourcing feasibility is visible while you’re still designing, so fewer ideas fall apart when you’re too far into the process. And nobody re-keys the same product into four systems, which sounds like a small thing and isn’t — ask any design team how much of their week goes on admin that only exists because the tools don’t talk to each other. We haven’t made anyone design faster. We’ve given them back hours they should never have been spending.

As AI moves deeper into retail, where do you see human creativity becoming more valuable, not less?
With AI, it’s become really easy for anyone to generate output. In fact, I believe we are drowning in a sea of AI-generated options, and what’s become even more important is taste. The ability to really look at an output and decide whether it should exist at all. Taste,unfortunately,cannot be outsourced to AI. It is something that you cultivate over many years of looking at things carefully. And this isn’t just limited to design; a marketer deciding what a brand stands for, a buyer who can predict changes in trends, all those become even more valuable in an AI world. AI lets all of them compress timelines and move from idea to production, but deciding what to make, for whom, and why still requires human ingenuity.