Glance, a Google-backed consumer tech company, launched Glance AI on Thursday. This new app allows users to try on AI-curated outfits using just a selfie. 

The app recommends fashion styles that suit each user’s look, body type, and preferences. Users can explore over 400 global brands and shop directly from the app.

“The internet is being rewritten from curated feeds to AI-generated realities…This is one of the most ambitious technological efforts in our journey and a bold step toward shaping the future of AI-driven consumer experiences,” Naveen Tewari, founder & CEO of InMobi and Glance, added.

Unlike traditional e-commerce apps, users don’t need to scroll through endless lists of products. Instead, Glance AI generates looks that are styled specifically for them and shows how they’d appear in real life using AI models. 

The company claims that 1.5 million users in the US have tried the app, and nearly half turned into repeat customers. So far, over 40 million style requests have been made, and 40% of users tap through to shop after seeing their looks, Glance said. 

The app gives people control over their data while offering an experience that feels personal and engaging.

Why not use in-house LLM’s?

In an interview with AIM, Mohit Saxena, co-founder and CTO of InMobi and Glance, stated that with foundational models like GPT, Claude, and Gemini developed over a decade of research and billions in investment, there’s little incentive to “reinvent the wheel.”

Thus, the company is leveraging these powerful models—sometimes open-source like DeepSeek and other commercial models like Gemini Pro and fine-tuning them for specific, localised use cases. 

Saxena mentioned that the real intellectual property (IP) lies in “data preparation, contextual layering, and post-processing.” As a result, this multi-model orchestration is likely to remain the company’s approach going forward. No single model today can master every step in the chain, Saxena added.

However, there are limitations. Most foundational models are predominantly trained on Western data, which results in a lack of understanding of Indian and broader Asian cultural nuances. 

This underscores the need for contextual AI tailored to India. “Localisation is the key. We need AI built for India and trained in our languages, dialects, and cultural contexts. That is the future I believe in,” Saxena added.

Bengaluru Team Behind the Scene

Glance AI’s proprietary tech stack was built entirely by the engineering team in Bengaluru. 

Running advanced AI models typically requires expensive GPUs. In the early stages, generating just 20 images could cost over $100, Saxena revealed. However, through “significant engineering effort,” the team optimised the process, dramatically reducing costs while maintaining high output quality.

Previously, creating a personalised style profile required around 20 images per user. A significant innovation changed that. The system now works with just a single selfie. 

While India remains an important market, its core business spans across the US, Europe, China, and other regions. 

“This is the first platform to be available simultaneously on the lock screen, ambient screens, and as a standalone app, offering a seamless and unique user experience across devices worldwide,” Saxena stated.
Glance AI has partnered with over 400 fashion brands, including Ajio, Zara, and ASOS, with plans to expand further. While the platform is launching focusing on fast fashion, it will soon branch into other categories such as sneakers, eyewear, cosmetics, makeup, lifestyle products, and home decor.

The post Glance Launches AI Shopping App to Enhance Consumer Experience appeared first on Analytics India Magazine.