What If You Could Search for Abstract Art the Way Spotify Searches for Music?
I spent thirty years building things that scaled. Then I started making things that couldn't.
Abstract Options!
In 2015, Spotify did something the music industry had never done. It didn't just catalog songs. It built a language for how music feels — and made that language searchable. Discover Weekly didn't ask you what genre you liked. It mapped where you lived in a vast dimensional space of sound, tempo, energy, and emotion, and found what was waiting for you just outside your known territory.
Nobody had done that for music before. Nobody has done it for art.
Not yet.
Think about how you currently find abstract art. You type "abstract painting" into a search bar and get ten thousand results organized by price and popularity. You walk into a gallery and someone explains why the work in front of you is significant. You browse Artsy or Saatchi and filter by color, by size, by style — as if a painting were a product with specifications.
None of these systems are built for you. They are built for inventory.
The fundamental problem is this: abstract art doesn't have a language. That's not a failure of abstraction — it's the point of it. A great abstract painting communicates something that exists before words, something that lives in the body before it reaches the mind. The moment you try to describe it as "gestural" or "meditative" or "warm-toned," you've already lost the thing itself.
And yet we keep trying to search for it with words. Because that's all the infrastructure gives us.
I've been thinking about this problem for a long time — first as a collector, then as a painter, and most recently as someone who spent thirty years building companies and couldn't stop seeing the gap.
I grew up in New Delhi, the son of an art critic who championed the great Indian modernists — Souza, Husain, Swaminathan. I absorbed a particular way of looking at abstraction, one that took it seriously as a philosophical practice, not just an aesthetic category. And when I returned to painting full time in 2020, after founding four companies, I found myself sitting in a studio in Phoenix with 252 paintings and no good way to help the right person find the right one.
Not because the paintings weren't findable. Because the language to find them didn't exist.
So I built it.
The Taxonomy of Abstraction is a nine-dimensional scoring framework for abstract painting. It measures Texture, Form, Colour Range, Palette, Mood, Space fit, Semantic Intent, Sentiment — and a dimension I call the Abstraction Quotient, which positions a work on the spectrum from pure representation to pure abstraction.
Every painting in my catalog is scored across all nine dimensions. When someone comes to the Abstraction Engine at rituart.com and describes a room, a feeling, a moment in their life — the system doesn't search keywords. It maps their context against the dimensional space of 252 paintings and finds the ones that belong together.
The results aren't recommendations in the way Netflix recommends. They're encounters. The painting that surfaces isn't the most popular or the most recently added. It's the one that has been waiting for exactly this person.
That's the Spotify insight applied to art: not cataloging paintings, but building a language for how paintings feel — and making that language searchable.
But here's what Spotify never had to reckon with, and what makes this problem harder and more interesting for art:
Abstract paintings have a 1-to-1 relationship with the viewer.
A song can be loved by fifty million people simultaneously. A painting — a specific abstract painting on a specific wall — belongs to one person. The encounter is singular. Which means the discovery system can't optimize for popularity or consensus. It has to map to an individual, to the specific shape of what they're building in their life, in their space, in their way of seeing.
Moods change. But a painting you've spent serious money on will live with you for a long time. It will change the room. It will change what you think in that room. The match has to run deeper than how you feel on a Tuesday afternoon.
I call this system ArtGraph.
The name matters because it describes the infrastructure, not the experience. A graph positions things in relationship to each other — paintings in relationship to dimensions, dimensions in relationship to collectors, collectors in relationship to encounters they haven't had yet. It's not a recommendation engine in the consumer sense. It's a discovery architecture.
The Abstraction Engine at rituart.com is the proof of concept — built on my own catalog, trained on real sentiment data from real viewers, already surfacing encounters that keyword search would never find. Rangoli in Permanence, my 5×5 foot acrylic on canvas, has registered the same emotional response — moves me — across three independent sessions from three different visitors who had never met. The painting found its people. The system recorded it.
That's not curation. That's signal.
The larger question — the one that keeps me in the studio and at the desk in equal measure — is what happens when this infrastructure exists at scale.
What if every abstract painter could score their work against the same framework? What if a collector in Tokyo and a designer in São Paulo and a first-time buyer in Phoenix could all navigate the same dimensional space, finding work that belongs to them specifically, from artists they would never have encountered otherwise?
What if Christie's didn't sort by provenance and price, but by how a work makes you feel in the specific room you're trying to fill?
What if the discovery language became the product?
Spotify didn't set out to replace music. It set out to make music findable in a way that felt inevitable once you experienced it. You forgot there had ever been another way.
That's what ArtGraph is for.
Not curation. Encounter.
ArtGraph is the working name for the dimensional discovery framework built on the Taxonomy of Abstraction. The proof of concept lives at rituart.com. The book is forthcoming.
