Virtual try-on startup Doji has announced a $14 million seed funding round, aiming to revolutionize online apparel shopping through engaging personalized avatars. Led by Thrive Capital and joined by Seven Seven Six Ventures, this round will enable Doji to refine and expand its proprietary AI models to enhance realism and user experience.
Founded last year by entrepreneurs Dorian Dargan and Jim Winkens, whose combined names inspired the startup’s identity, Doji leverages cutting-edge diffusion-based artificial intelligence technology. Users can create lifelike avatars by uploading six facial selfies and two full-body images. These highly realistic digital representations then help shoppers explore and evaluate fashion offerings tailored to their personal style and favored brands.
Dargan brings extensive experience in digital avatars and interactive design from prior roles at Apple, where he contributed to VisionOS, and Meta, where he developed games and immersive experiences for Oculus Quest. Winkens, meanwhile, honed his expertise in generative AI as a researcher at DeepMind and Google.
The idea for Doji originated when Dargan and Winkens connected online in 2022 and noticed the viral popularity of AI-driven avatar creation platforms such as Lensa. Recognizing users’ emotional attachment to personalized avatars inspired them to imagine a similar experience adapted specifically for fashion, built around photorealistic virtual try-ons.
Early reactions to Doji have been enthusiastic, with initial users drawn to the enjoyable, social aspects of experimenting with new looks through personalized avatars. Thrive Capital’s Miles Grimshaw said he was captivated by the potential of Doji from his first experience, referring to the app as a “virtual mirror” that makes online shopping more intuitive and enjoyable. According to Grimshaw, Doji’s approach uniquely integrates users at the heart of every interaction, offering an accessible, engaging alternative to the conventional browsing experience.
Currently, Doji’s app remains invitation-only, available in over 80 countries, with the avatar creation process taking approximately 30 minutes. While existing virtual try-on solutions typically rely upon limited data inputs, Doji opted for a more comprehensive setup to enhance avatar quality and accurately capture users’ distinct features. The app allows users not only to preview clothes available directly within the platform, but also to check styles from external retailers by simply submitting product links.
Although early testing has surfaced minor issues with avatar realism, such as inaccuracies around height or body shape, Doji offers users the option to retrain avatars using new photos to ensure greater authenticity. In the future, Doji plans updates to improve clothing fit predictions and reduce avatar creation times significantly. Additionally, the startup intends to integrate direct purchasing capabilities within the app, removing the friction of redirecting users to external shopping sites.
Dargan believes Doji’s core strength lies in its team’s deep technical and design experience, specifically their careful attention to visual quality and user perception. Unlike predecessors who may have lacked either technical capabilities or design sensibility required for widespread adoption, Doji positions itself uniquely at the intersection of advanced AI and thoughtful, human-centered design.
Ultimately, Doji aims not just to simplify the process of discovering fashion online but to transform it into a dynamic, inherently social experience—making virtual try-ons both enjoyable and compelling for users worldwide.
