Despite advances in e-commerce personalisation, physical stores remain largely disconnected from customer data. The long-discussed vision of omnichannel retail—where stores recognise customers and personalise experiences—remains unrealised largely due to organisational silos.
“When you shop online, your footprint is there. But when you walk into a store and leave, brands don’t really know you,” Sreeraman Mohan Girija, Founder of retail tech solutions company Fynd, tells AIM.
A fundamental mismatch between retailer product descriptions and consumer search behaviour is driving billions in lost sales and missed commercial opportunities. According to a recent GS1 India study, inconsistent, incomplete, and inaccurate product information is leading to approximately ₹5,000 crore ($530 million) in annual revenue losses across India’s e-commerce and quick commerce landscape.
For retail technology firms like Fynd building AI solutions, success depends less on raw model capability and more on contextual adaptation, such as adding vernacular and voice AI integration to cater to India’s linguistic diversity. “We almost have to build a layer on top of existing solutions,” Girija remarks.
The company’s store operating system attempts to bridge this gap by integrating purchase history, CRM data, and in-store interactions into a single platform.
“Retailers globally are dealing with fragmented legacy systems and rising expectations for seamless omnichannel experiences. Building a platform that can unify online, in-store, and backend operations while remaining modular, scalable, and quick to deploy has been a complex but critical challenge,” Girija explains.
Foundations of Retail Tech
Fynd was founded in 2012 by Girija, Farooq Adam, and Harsh Shah under the name Shopsense Retail Technologies. Girija and Adam, both IIT Bombay alums, conceived the idea at the institution and bootstrapped it with around $30,000 by winning business plan competitions.
The Mumbai-based company aimed to digitise offline retail through in-store solutions, but the founders soon identified a larger opportunity in connecting fragmented store inventory to online consumers. This insight led to a pivot in 2015, rebranding the company as Fynd and launching a fashion e-commerce platform that bridged online and offline retail.
Its early innovation involved installing large touchscreens at retail stores, allowing shoppers to mix and match outfits, and making offline shopping more interactive.
“From there, we pivoted to a B2C marketplace model, positioning ourselves alongside platforms like Amazon and Flipkart. This phase was built on strong inventory integrations with retail partners. A key differentiator was our ability to bridge geographic gaps, allowing customers in tier-2,3 cities to shop seamlessly from premium stores in Mumbai and Delhi,” he adds.
Over time, the company expanded beyond in-store engagement into a full retail technology stack covering website creation, warehouse management, transport systems, and transaction infrastructure. It raised over ₹100 crore from over 50 investors, including Google, IIFL, and Kae Capital, before being acquired by Reliance Industries (RIL) in 2019 for ₹295 crore (approx $40 million), as part of its push for online-to-offline (O2O) commerce.
From O2O to B2B
Since becoming part of RIL, the business transitioned into a pure B2B retail technology platform. Today, its offerings span digital storefronts, marketplace connectivity through Fynd Konnect, POS, and in-store technology via StoreOS, and end-to-end supply chain solutions across order, warehouse, and transport management systems.
AI now sits at the centre of nearly every product it offers.
“We’ve built a strong AI-led portfolio with products such as Kaily, Pixelbin, Boltic, and GlamAR. This evolution is further strengthened by capabilities like AI-powered PIM and AI fashion design (Fynd Create),” Girija notes.
It serves over 20,000+ stores and 300+ enterprise retailers, including in the UK and European markets and prominent clients like the Landmark Group, PUMA, and Style Baazar, offering a modular commerce stack that unifies in-store, online, and logistics operations.
Fynd’s revenue is powered by three commerce products: Storefront, which helps brands sell online at scale; Store OS, which brings intelligence to physical retail; and Konnect, which connects brands to every major marketplace from a single platform. Meanwhile, Fynd Create’s AI-powered solution acts as an end-to-end creative studio, helping brands, retailers, and designers digitise fashion development and speed up cataloguing.
The startup earned around ₹200 crore (approx $24 million) in FY25.
Why Indian Retail Breaks Standard AI Systems
While global AI models provide powerful foundations, deploying them in India reveals limitations rarely encountered in more standardised markets such as the UK or Europe.
A simple product search illustrates the problem.
“The same product can be known by different names in different local languages,” he explains.
To address this, engineers must build additional intelligence layers that interpret intent rather than literal input. These include synonym and language mapping layers, regional intent recognition, attribute normalisation systems, and others.
Girija states that one of the toughest challenges has been scaling a unified commerce platform across diverse markets with very different regulatory, operational, and consumer requirements.
“As we expand into regions like the GCC (Gulf Cooperation Council), we have had to localise deeply from compliance frameworks like e-invoicing, tax structures, or even language and storefront experiences,” he says.
At the same time, retailers globally are dealing with fragmented legacy systems and rising expectations for seamless omnichannel experiences. Building a platform that can unify online, in-store, and backend operations while remaining modular, scalable, and quick to deploy has been a complex but critical challenge.
“From an organisation-building perspective, a large chunk of our growth happened during the COVID-19 remote work years. That period taught us a lot about building an organisation that can withstand testing periods and emerge stronger,” Girija recalls.
The complexity of Indian retail data is turning the country into an unexpected testing ground for Fynd’s AI systems.
India does a lot of region-specific innovations, which is not a struggle in the UK market.
The reason lies in data standardisation. European retail ecosystems operate with relatively uniform naming conventions and consumer behaviour patterns.
“In India, the customer talks in one way, the brand uploads the catalogue in another way,” Girija quips. Solutions that succeed in India, however, tend to scale globally with fewer modifications.
One of the biggest shifts underway is the rise of conversational commerce. However, for India, voice-based commerce introduces another layer of complexity. India’s dialect diversity presents challenges even for advanced speech models.
Interestingly, Fynd’s early deployments show that users often communicate through voice messages rather than real-time conversations. “People are not chatting live,” he observes. “They are essentially giving voice messages, just like WhatsApp.”
The company is also exploring what it calls “autonomous commerce”—a future where AI enables individuals to launch and run retail brands independently.
Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.