The OMNIOUS.AI team had the privilege of participating in AWS re:Invent 2024, held in Las Vegas from December 2 to 6. This was the 13th annual event for re:Invent, a global hub for the latest in cloud computing, AI, and emerging technologies.
This wasn’t just another conference—it was an opportunity to showcase Vella, our virtual try-on service, alongside some of the biggest names in the generative AI industry.
In this post, we’ll share insights from the event, including key takeaways from the sessions as well as highlights from our experience showcasing Vella to hundreds of attendees at the EXPO hall.
Key Takeaways from the Sessions
Our team attended multiple sessions, gaining insights into the latest trends in cloud-based AI and AWS technology. Here’s what stood out:
AWS re:Invent 2024 offered a clear view of where cloud-based AI is heading. The rapid evolution of Amazon Bedrock's ecosystem stood out as particularly relevant for Vella's future development.
A defining moment came during Amazon CEO Andy Jassy's keynote appearance, where his emphasis on 'CHOICE' in the AI landscape struck a chord with our vision. His message that different use cases demand different AI solutions aligns perfectly with our approach at Vella, as we carefully evaluate new technologies while maintaining the excellence of our core virtual try-on experience.
The expansion of Amazon Bedrock with new foundation models (Nova, Canvas, and Reel) and partnerships with specialists like Luma AI for video generation demonstrates the maturing ecosystem for advanced AI applications. The new features in model distillation and multi-agent collaboration present opportunities to optimize our AI infrastructure while maintaining quality. Additionally, the introduction of the Amazon Bedrock marketplace opens exciting possibilities for offering our specialized fashion AI models to a global audience.
These developments have both validated our current direction and revealed new opportunities for Vella's growth, all while maintaining our focus on delivering exceptional virtual try-on experiences.
Showcasing Vella at the EXPO Hall
Throughout the week, our booth visitors were able to experience a hands-on demonstration of the Vella virtual try-on and see firsthand how our AI adapts to individual image styles and delivers authentic results.
One memorable moment came when an AI/ML consultant from a Big 4 firm remarked:
"I’ve seen many generative AI tools, but this [Vella] is by far one of the best!"
One common question we heard was:
“What model is Vella based on?”
Vella is the advanced commercial version of the IDM-VTON model which we introduced in our research paper “Improving Diffusion Models for Authentic Virtual Try-on in the Wild.”
One notable suggestion came from a VP at an email marketing and automation company. He envisioned a unique use case:
"Retail clients could send personalized emails to customers, including generated images of those customers wearing clothing items they added to their carts but hadn’t yet purchased."
This feedback highlighted how Vella could transform customer engagement by creating a seamless, personalized shopping experience.
Looking Ahead
AWS re:Invent 2024 was more than just a conference—it was a celebration of innovation, collaboration, and learning. From engaging conversations at the EXPO hall to game-changing insights from industry leaders, the event reinforced our commitment to pushing boundaries in fashion tech.
As we look ahead to 2025, we’re excited to explore new opportunities, continue refining Vella, and share our journey with the world.
Want to see what Vella can do? Try it out here or follow us on Instagram and X for the latest updates.