Adobe’s Varun Parmar Redefines How Iconic Brands Are Using AI at Scale

For years, Adobe has been the gold standard for photo and video software in the business market. But today, it faces a very different competition. A wave of AI-powered startups can now generate and edit content in seconds, often at a fraction of the cost. Tools like Higgsfield promise fast, click-to-video workflows for marketers who value speed above all else, while platforms like Canva have made “good enough” design accessible to almost anyone.
Adobe’s answer to that disruption isn’t to produce for everyone. It’s a focus on something more difficult: orchestration. The idea is to move content from concept to delivery without breaking product rules, budget or trust. And the person leading that strategy is Varun Parmar, Adobe’s general manager of Firefly Enterprise and GenStudio, the company’s main AI platforms launched in 2023.
Firefly is Adobe’s family of generative AI models, powerful image, video, audio and vector models across the Creative Cloud. GenStudio is a business layer, designed to connect those capabilities directly to marketing programs like Google and Amazon Ads.
“As brands compete for consumer attention across fast-moving platforms including social media, mobile apps and e-commerce, the demand and need for relevant, outstanding content is on the rise,” he told the Observer.
The creative industry is no longer short on tools that can produce images or copy. What’s missing is a way to make all of that output usable within large organizations, where content requires approval, formal testing, version control and product consistency. “Brands are looking for a technology partner with a deep understanding of their creative and marketing flow,” says Parmar.
This is where Adobe believes it still has a structural advantage. Long before generative AI entered the picture, Adobe built tools to manage complexity: Workspace for project management, Knowledge Manager tools for content libraries, and Creative Cloud for productivity. Firefly builds on that foundation.
Parmar knows this world well. He started at Adobe in the early 2000s, working on products like Acrobat and Document Cloud, then left to spend time at startups, including roles at Miro and Box. He returned in 2024 to lead Adobe’s AI push. The bet is that while AI can generate infinite content, brands don’t actually want infinite. They want manageable, product content that can grow without turning into chaos.
Adobe’s approach to AI in action
One of the clearest examples of Adobe’s strategy plays out at Serta Simmons Bedding, a century-old mattress company with an incredibly sophisticated marketing mission. Serta sells directly to consumers, through wholesalers and through thousands of retail outlets. Each channel needs different versions of the same campaign.
That turns content into a volume problem. A single campaign can require hundreds of variations: different formats, sizes, languages, and vendor-specific versions. When those assets sit in disconnected systems, costs rise and teams slow down.
Serta embraces GenStudio not as another tool, but as the core of a new creative work model. The creative process now starts with a structured input that matches the product, channel and format, so that the generative AI connects to a defined pipeline instead of creating a random output.
Designers still work within Creative Cloud. Firefly services are embedded directly into production, handling tasks such as resizing, background generation and format adaptation, all within product guidelines.
“Firefly APIs can quickly generate product-safe background options, and Photoshop APIs integrate banner ads for each channel, resulting in a 50 percent reduction in digital content production hours. All this is done while maintaining product guidelines, which have been used to train the AI model,” said Parmar.
What makes a system work is orchestration. Assets flow back to shared systems. Updates happen in the same place. Instead of multiple tools bundled together, Serta now has a single virtual content pipeline.
“AI gives marketers speed and efficiency, but the real opportunity is to use it thoughtfully…Instead of replacing creativity, AI has made us more productive and expanded the creativity possible,” Tim Oakhill, chief marketing officer at Serta Simmons Bedding, told the Observer. “AI (in SSB) is used as an efficient tool, not a consumer-facing human creation tool, which allows our teams to move quickly without losing emotional depth.”
Adobe says that GenStudio has helped Serta increase the number of customized products it produces tenfold. That kind of scale is shrinking. According to Adobe research, 71 percent of marketers expect the demand for content to increase fivefold by 2027.
Serta is not the only product that uses Adobe in this way. Tapestry, owned by Coach and Kate Spade, uses Firefly Custom Models trained in proprietary handbag imagery to produce brand-specific designs and marketing content.
Newell Brands, parent company of Elmer’s Glue, Sharpie and Paper Mate, uses Firefly to accelerate graphics and visual variations across its product lines.
Mattel uses Firefly to package and tell stories around Barbie. Coca-Cola worked with Adobe to develop Firefly Design Intelligence, an AI-powered design system aimed at keeping global brand campaigns visually consistent across markets.
What all these companies have in common is the desire to eliminate friction: fewer repetitive tasks, fewer manual changes and fewer lost materials.
“Our brand strategy, with productive AI, is about freeing up time and mental space for marketers and designers. It’s about managing and deploying AI responsibly,” Parmar said. “That’s been our focus from the beginning, from Adobe Firefly’s commercial-safe models and content insights, to working with brands on data management and ensuring content is ethical and avoids bias.

!function(f,b,e,v,n,t,s)
{if(f.fbq)return;n=f.fbq=function(){n.callMethod?
n.callMethod.apply(n,arguments):n.queue.push(arguments)};
if(!f._fbq)f._fbq=n;n.push=n;n.loaded=!0;n.version=’2.0′;
n.queue=[];t=b.createElement(e);t.async=!0;
t.src=v;s=b.getElementsByTagName(e)[0];
s.parentNode.insertBefore(t,s)}(window, document,’script’,
‘
fbq(‘init’, ‘618909876214345’);
fbq(‘track’, ‘PageView’);



