• Bengaluru's Ayna uses proprietary AI models to generate studio-quality product photos at scale.
  • Here's how the Zomato Mafia startup is changing Indian e-commerce.

Indian E-Commerce Has a Photography Problem. Ayna Just Got Funded to Fix It.

Running a fashion or home décor brand in India means one thing is always eating your budget and your calendar: the photoshoot. Booking a studio, hiring models, coordinating stylists, shooting hundreds of SKUs, retouching, reformatting for different platforms — the average mid-size D2C brand spends between ₹50,000 and ₹5 lakh per photoshoot, and still ends up with images that look inconsistent across their catalogue. Scale up to thousands of products and the problem compounds fast.

Bengaluru-based Ayna was built to eliminate that process entirely. Founded in 2023 by Aastha Rajpal and Yash Bansal, Ayna enables brands to produce studio-quality product photoshoots at scale without the need for physical photoshoots. In June 2024 the startup closed $1.5 million in seed funding led by Inflexor Ventures, with participation from four institutional investors and sixteen angels — a broad syndicate for a pre-revenue company that reflects how seriously India's early-stage investor community is taking generative AI in commerce.

The Founder Who Shut One Startup Down to Build a Better One

Yash Bansal's path to Ayna is the kind of founder story that doesn't get told often enough. He previously built Pet Perfect, a pet care app that won Best App recognition on Google Play in 2022. Then he shut it down.

"Just last year, I faced the bittersweet task of winding up operations at Pet Perfect, a venture into which I had poured my heart and soul," Bansal wrote on LinkedIn when announcing Ayna's funding. "It was a tough decision, but sometimes growth requires letting go. Soon after, I embarked on a new journey at Zomato, where I learned and grew immensely. But deep down, my heart belonged to the world of startups and AI."

That Zomato stint matters more than it might seem. Bansal is part of what Inc42 has documented as the "Zomato Mafia" — the alumni network of former Zomato employees who have gone on to found startups. By 2024, 61 ex-Zomato employees had launched 51 startups, a network that carries both credibility and warm investor relationships for anyone emerging from it. Bansal brought co-founder Aastha Rajpal, who leads product and technology, and together they built Ayna's core AI infrastructure before raising external capital.

Further validation came from an unexpected direction: Ayna was selected for Stability AI's Stable Founders Program, recognising it among the top 50 startups globally working with foundation models — a signal that the underlying technical approach was credible beyond just its commercial traction.

How the Technology Actually Works

Most AI image generation tools available to brands are built on top of general-purpose diffusion models — Stable Diffusion, Midjourney, DALL-E — fine-tuned lightly for product use cases. The outputs are impressive but inconsistent: a product might look right in one image and slightly distorted in another, colours shift, proportions drift.

Ayna's proprietary Compound Foundational Models (CFMs) give brands granular control over every detail of their product images in a way that generic tools cannot match. The platform allows a brand to specify exact requirements — set the background, adjust lighting and mood, define a brand colour palette — and then generate consistent imagery across an entire catalogue.

The virtual model capability is where Ayna pushes furthest. The platform can create and customise an entire virtual human model around the clothes or any product, eliminating the necessity of human models for photoshoots. Its GenAI-based platform allows creation of virtual human models with multiple options across ethnicities such as American, Russian, Indian, and Indian sub-cultures such as Kashmiri, Bengali, and others. It also gives options to choose different body types, gender, age, as well as eye and hair colours.

For an Indian fashion brand selling to a diverse domestic market, this matters specifically. A brand selling to consumers in Kerala, Punjab, and Maharashtra cannot assume one model presentation resonates with all three audiences. Ayna lets them generate localised imagery at effectively zero marginal cost per variant — something that would require multiple shooting days with multiple model bookings in the traditional workflow.

The Clients, the Scale Ambition, and What Comes Next

At the time of the funding announcement, Ayna's client roster included Reliance Retail's Clovia, WomanLikeU, Bummer, and D2C furniture brand Wakefit. The Wakefit addition was notable — it represented Ayna moving from pure fashion into home décor, a category with its own acute photography problem: furniture and décor simply cannot be photographed efficiently at catalogue scale without enormous studio and logistics overhead.

Ayna was looking to grow its customer base to over 1,000 by the end of 2024 from 15 at the time of funding. Whether that target was hit is unclear, but Tracxn data shows 27 employees as of July 2025 — a team that has grown meaningfully since the two-founder founding stage, suggesting operational scaling is underway.

The go-to-market expansion has two tracks running in parallel. Ayna is also creating a portal for long-tail customers where any user can enter and do photoshoots, moving beyond enterprise-only deals toward a self-serve model that mirrors how Canva opened design to non-designers. Simultaneously, international expansion — particularly into US fashion markets and European hubs — is on the roadmap. India remains the proving ground, but the unit economics of AI-generated photography work equally well in any geography where studio costs are high.

Bottom Line

Ayna sits at the intersection of two fast-moving curves: generative AI maturing to the point of commercial reliability, and Indian D2C brands hitting a scale inflection where photography costs become a genuine operational constraint. The $1.5 million seed is small relative to the problem size, but the CFM technology stack, the early enterprise validation from brands like Clovia and Wakefit, and a founding team with Zomato DNA and Stability AI recognition make this one of the more technically credible generative AI startups to emerge from Bengaluru's 2023–24 cohort. The next milestone to watch: whether the self-serve platform materialises, and whether the US market entry adds the revenue base needed for a Series A.

Edited by Nabarun