Company: Omninity.ai * Role: Product & Brand designer * Product: Website, platform & branding
Omninity is a fintech startup building AI-powered agents, designed to turn business workflows into seamless, human-centred dialogue. As Senior Product & Brand Designer, I led end-to-end design of the brand identity, the creation of the website and UX for Omninity’s platform.
Working closely with the founder, I established a unified visual system and interaction experience that positioned the product as both intelligent and trustworthy: differentiating in a crowded AI landscape and delivering an experience that both enterprise buyers and end users could trust. Design constraints included a tight timeline, limited internal design resources, and the need to align both brand and UX with evolving product architecture.
A website and platform fit for the future of AI
AI-first fintech brands often leaned too far into complex UX and futuristic aesthetics, risking alienation of less technical stakeholders. Based on this, I established two foundational design imperatives: to make intelligence feel human and to make complexity feel transparent.
The UX challenge was to reduce cognitive load and make the conversation flow feel intuitive, even as complex financial rules and data logic happened behind the scenes. I designed the feeling around three key principles: modular conversational UI, progressive disclosure of data insights, and human-in-the-loop handoff transparency.
Post-launch, Omninity achieved meaningful traction: the unique brand identity and website helped secure early-stage client interest and positioned it credibly in enterprise sales conversations.
Internally, the unified brand system has provided a foundation for all subsequent product modules, allowing the team to scale consistently (building durable design assets for future growth).
This project reinforced for me the power of aligning brand and product design in tandem. Key lessons include: the importance of making hidden logic visible (especially with AI), and designing for human-like interactions even when the underlying system is novel.
