Serendipity Engines & ChromaverseSerendipity Engines

A bold reimagination of how AI concierges and fluid interfaces enhance the thrill of how people search, discover and buy.

At dotdotdash, we believe AI shouldn’t replace human curiosity in commerce. We believe AI is a serendipity engine fueling the feeling of the search, not erasing it. It serves as a system that fosters unexpected, valuable, chance connections and discoveries by encouraging curious exploration over just goal-driven searches and deterministic outcomes. That’s why we built Chromaverse – an AI concierge prototype that helps people shop by color. It scans catalogs by a shopper’s color preferences, and surfaces relevant products and information through a fluid interface that adapts to the context of their needs. It goes beyond the traditional e-commerce experience. It’s an experiment in AI-driven commerce that isn’t focused on convenience and ease, but rather ignites the feeling of serendipity and surprise that people love to get from a successful shopping experience.

As generative AI and LLMs enter e-commerce, there’s a common assumption by both businesses and consumers that AI should deliver precise, deterministic answers to queries. That the outputs are the same every time from the same input. But these models are inherently probabilistic — they generate outputs based on patterns, not fixed rules. That means the same input can yield different results, which poses challenges in contexts where accuracy, consistency and reliability matter. For brands that want to maintain a tightly controlled grip, this unpredictability can feel like too great a risk of eroding consumer trust. As a result, many e-commerce experiences are either hyper-optimizing and hyper-homogenizing AI to eliminate variability, or avoiding the use of it altogether.

Generative AI’s probabilistic nature should be viewed as its strength, not its liability. It's not about giving the answer. It's about offering a possibility that may not be expected. Generative AI is great at randomness, surprise and discovery, eliciting the feeling, “I didn’t know I wanted that.” It is the perfect serendipity engine – a tool not to remove but to inject more emotional exploration and brand distinction into one’s search. For brands that want to maintain a tightly controlled grip, this unpredictability can feel like too great a risk of eroding consumer trust. As a result, many e-commerce experiences are either hyper-optimizing and hyper-homogenizing AI to eliminate variability, or avoiding the use of it altogether.

The design principles of our prototype imagine AI and e-commerce beyond hyper-optimized homogeneity and conventional chatbots, toward a more distinct, future paradigm.

Instead of typical color filters, Chromaverse transforms a simple color query into an open journey – exploring the interplay between how shades and styles mix, match and clash.

Chromaverse blends visual browsing, conversational UI and PDPs into one fluid, responsive interface. It adapts to a person’s intent in a nonlinear journey — offering possibilities without breaking context or sending them elsewhere. The corner avatar shifts shape and color based on their actions, while the nav bar toggles between styles and items that match their choices.

Chromaverse is tailored to a person’s taste, surfacing the perfect shades and styles. Guide it through color chip selection, text prompts and real-world hue samples captured by their phone camera.

Chromaverse offers subtle, real-time nudges to guide a person’s curiosity and inspire trust — from color theory tips to deeper stories behind each clothing item, going far beyond the typical constraints of a PDP.

Chromaverse isn’t a broad generalist or an over-specialized tool. It’s T-shaped — trained on deep color theory and curated fashion archives, yet versatile enough to meet a person’s needs.

Chromaverse invites, but it always keeps a person in charge – whether they choose to stay on their original path, or wander into new unexpected side quests.