Brand & Product Design
Professional Work
June 2022 - April 2023
Building a generative AI product that enables brands to create diverse, inclusive content at scale, tackling both the technical challenge of creative alignment and the social mission of authentic representation.
Introduction
Branded content lies in the last 10%
The pictures and videos we see in consumer advertising today reflect a small population of our rich ethnic diversity. Brands need to tell compelling stories daily to stay relevant and top-of-mind. However, lean teams often lack the resources to do so, such as professional photoshoots. As a result, they are left with stock photography, which can become repetitive for their audience.
While stock photography covers 80-90% of content needs, it's the final 10% that determines whether content truly feels branded. This missing piece is incorporating brand colors, specific photography styles, and authentic visual aesthetics, which are the hardest to achieve but makes all the difference. Combined with inappropriate representation of the target market, creatives fall flat and are ignored.
Real world problem
Despite numerous companies working on cancer treatment in the region, there was a striking lack of photography representing actual cancer patients. The same stock photo appeared across multiple companies' materials: a clear disconnect between the diverse patient population and how they were visually represented.
This pattern extends globally. Traveling to different countries revealed how advertising consistently failed to reflect local demographic diversity. This misrepresentation highlight a systemic problem: the creative industry lacked tools to authentically represent the communities they served.
Overusage of a stock photo across multiple entities for multiple markets
The Opportunity
Mature generative AI technology with a mission-driven approach to authentic representation could bridge this gap between generic stock photography and personalized brand content.
About Re:cast AI
Re:cast AI enables anyone to achieve a new, diverse, and inclusive voice in their content.
We believe content can inspire and inform. Re:cast AI provides a scalable, transformational content creation platform that enables brands to effortlessly bring greater inclusivity and diversity to our world.
This technology goes beyond productivity. Consumers want to see themselves in the ads they are served. It is essential that all demographics and ethnicities are represented accurately, not just based on outdated biases and stereotypes. This is our overarching purpose. Our priorities are driven by the question: "How do we accurately and fairly represent the demographics around us?"
Brand Identity and Design
Re:cast AI Brand Guide
The brand identity emerged from our core mission of authentic human representation. We centered our color palette around skin-tone-inspired hues create a grounded, natural feeling that spoke to our "brand for people" philosophy to create a visual that felt approachable and inclusive.
The tensor symbol was chosen for as the recurring motif for its versatile symbolism. At its core, it represents multiplication, which was the problem we wanted to solve. Decomposing the symbol, there is a balance of assertiveness (the cross element) with harmony (the circle), and versatility to extract new shapes, which was symbolic of what the diversity we wanted to bring.
Product Design Evolution
Version 1
Proof of Concept
How would the user journey look like to create a new picture?
Starting with a basic text field for image description, similar to DALLE2's prompt box approach, we aimed to prove the capability existed for photorealistic, marketing-appropriate images.
Version 2
Creative alignment
How can we make it easy for people to create images without complex prompts?
While my co-founder tackled creative alignment from the AI technology side, I focused on the product design challenge: moving from an intimidating empty text field to familiar form-style inputs:
Description box
Whotography style selection
Brand colors definition
Aspect ratio and spacing controls
We wanted to give users structured, intuitive ways to express what they wanted without needing to master prompt engineering.
Version 3
Workplace Use
How can we integrate this seamlessly into how marketing teams work?
Rather than just proving capability or improving generation quality, wanted the app to feel native to existing creative workflows while enabling the collaborative aspects that teams naturally needed for maintaining consistent brand assets.
Team-shared brand style libraries
Ability to upload brand guidelines and automatically extract settings
Subscription-based access to generated images and defined brand styles
demo
Collaborative dynamic
The difference in background between my co-founder as I (his in AI/technical product development, mine in marketing) created productive tension throughout the product development process. While I focused on user needs, practical workflows, and ensuring the product solved real business problems, he pushed the boundaries of what was technically possible with AI generation. This collaboration was essential for tackling the creative alignment challenge from both sides: he improved the AI's ability to interpret and generate, while I redesigned how users communicated their vision to the system.
Though sometimes contentious due to our different technical expertise levels, this partnership ensured we weren't just building impressive technology, but creating a useful product that addressed genuine market needs.
Gratitude and Acknowledgment
Co-founder
Thank you for giving me the opportunity to work alongside this vision you had and help define the impact we can have in this world. Thank you for the complete trust you had in me to lead the marketing, brand, and growth aspects of the company. It was both humbling and empowering. Your patience throughout this journey taught me what actual product development looks like. I came out a better person and for that, I am eternally grateful.
Our Advisors
Our journey was enriched by advisors and friends who joined us at different stages, generously sharing their expertise and time. Product designers offered insights on key design considerations we hadn't yet explored. Growth experts provided guidance on scaling and go-to-market strategies. Their willingness to give us honest feedback, share their knowledge, and believe in our mission was invaluable. The time and energy they invested in helping us iterate and improve made a profound difference in our development as founders and as a product.
Reflections
August 2025 (2.5 years later)
There was so much heartache and tears at the end of this because we were on to something great, but we didn’t play the startup game well enough. I was able to pour all my heart and soul into the design of the brand and product, to obsess over the reason why we should exist, why anyone should care about us, and what we can actually do for people throughout the platform. The promise of AI, then and now, remains in the unimaginative productivity sphere - faster, better, cheaper - when there are greater problems that this technology can solve: e, climate, and society.
Today, most AI-generated images still carry that telltale "airbrushed" quality- perfect, polished, idealistic in a way that feels artificial. While the technology has advanced significantly, the creative alignment problem persists. Users still struggle to quickly translate their mental vision into generated reality.
Moreover, the storytelling element remains largely missing from current AI-generated imagery. The expressions look exaggerated, the scenarios feel constructed rather than captured. The greatest photographs used in creative and advertising tell stories. Until we solve for authentic narrative in AI imagery, we're building impressive technology without true creative impact.
Looking back at the problems we anticipated and the ambitions we had, it was definitely a shame that we never got far enough to implement the vision for an inclusive advertising mix and the platform that could solve it- something, I think, we still need today. I think it’s why it took me this long to share the work here.