Pack Futures: AI‑Driven Innovation with Ai Palette’s Som GanChoudhuri
Welcome to Pack Futures by 3D Color. Hosted by Jason Hauer and fellow 3D Color leaders, this series explores how leaders in CPG and retail harness innovation, technology and emerging consumer insights to shape the future of packaging.
In this conversation, we speak with Somsubhra GanChoudhuri, Co‑founder and CEO of Ai Palette. We discuss how Ai Palette’s platform analyzes vast amounts of consumer data globally to identify emerging trends and empower brands to launch products and packaging that truly resonate. We also dive into Som’s journey from Givaudan to entrepreneurship, highlighting examples from top CPG brands.
Pack Futures Interview with Som GanChoudhuri
Jason: For leaders in packaging, R&D, and design, what does Ai Palette do, and how is it different from other AI‑driven insight platforms?
Som: Ai Palette helps CPG companies create products and packaging that consumers will love. We have built what we believe is the world’s largest consumer data lake for our space, covering 28 countries across food and beverage, beauty and personal care, and now nutraceuticals and pet care. The platform analyzes public, real‑world signals in the local languages of each market, surfaces emerging needs before they become mainstream, and then turns those signals into action.
What makes us different is that we go beyond spotting trends. Teams move from discovery to creation and prioritization inside one workflow: identify opportunities, generate on‑brief concepts with Concept Genie, and quickly rank the strongest directions with Screen Winner for a specific country, channel, or occasion. The same pipeline informs packaging choices, including on‑pack language, color and imagery, structure, and sustainability cues that will resonate in market.
Jason: You spent time at Givaudan before launching Ai Palette. How did that experience shape what you built?
Som: At Givaudan I worked with market leaders and regional giants and saw a consistent issue. Traditional research told teams what consumers wanted today, but launches often reached the shelf one to two years later. By then tastes had moved. In 2018 there was no reliable way to forecast where preferences were heading, so my Co‑founder and I decided to build Ai Palette around predictive intelligence for tomorrow’s demand.
Jason: Let’s go deeper on your Foresight Engine. How does it work, and how does it inform packaging decisions specifically?
Som: Our Foresight Engine is designed to help you know what you do not yet know. It continuously reads real‑world signals from search, social, menus, recipe hubs, and e‑commerce to understand what people are actively seeking, trying, and buying, as well as what innovators are introducing.
We model those signals to reveal patterns across flavors, ingredients, occasions, positionings, and benefits, then tie them to packaging cues by market. That includes how consumers connect colors and imagery to emotions and seasons, which materials and sustainability attributes are rising fastest, and which formats or structures reinforce a product’s promise.
The output is not a generic trend list. It is market‑specific guidance a packaging team can use to choose a colorway for a flavor, dial in on‑pack language and descriptors, evaluate pack formats side by side, and prioritize concepts likely to win on a particular shelf.
Jason: How does this translate into tangible packaging and product outcomes? Any examples you can share?
Som: A leading beverage brand in the Philippines used our platform to design a limited‑edition holiday pack. We pinpointed color and imagery cues most associated with Christmas and New Year in that market, which helped the team build a package that exceeded seasonal sales targets.
With KitKat in the United States, the team screened dozens of flavor ideas in an afternoon and selected Blueberry Muffin for a limited‑time launch. It over‑delivered and ultimately became part of the ongoing portfolio.
We also helped Dole reposition pomegranate in the Philippines by shifting on‑pack emphasis toward vitamins rather than antioxidants, which unlocked meaningful growth. Another case is a global whiskey brand in China. By analyzing Chinese‑language data, we isolated the descriptors that signal premium and translated those cues into both messaging and on‑pack language.
Jason: You often talk about packaging as both aesthetic and functional. How is AI helping in each dimension?
Som: On the aesthetic side, generative tools let a design team visualize hundreds of plausible routes in hours based on flavor, ingredient, and positioning. It is not the final artwork. It gets you to the top four or five directions that merit further development and consumer testing. You can also rough out different structures and formats to see how a concept stands out at shelf and in a digital storefront.
On the functional side, simulation is becoming practical. One common question is shelf life. By modeling how formulation variables interact with packaging materials, barrier layers, and even color, you can get much closer to the target before formal stability testing. The goal is not to replace lab work. It is to cut iteration time, reduce upstream waste, and start pilots with a near‑right prototype.
Jason: How do you see AI changing the innovation and packaging process overall?
Som: Downstream steps still need people and labs. You will still do sensory, pilot runs, and line trials. The transformation is upstream. Predictive insight and generative tools compress the time it takes to discover opportunities, create on‑brief concepts, visualize packs, and prioritize winners. What took months can happen in a day for early screening, which reduces risk and increases the odds that your packaging and product reflect where the consumer is going.
Jason: Many leaders run AI pilots that never scale. What advice do you have for getting from experiment to enterprise impact?
Som: Start with an enterprise AI strategy and clear, measurable outcomes by domain. Many pilots are championed by passionate individuals but sit outside core priorities, which is why they stall. The companies seeing results pick a few functions where AI can drive ROI, then align cross‑functional processes and incentives. Leaders like those we hosted in Chicago tend to be innovators or early adopters. They test, learn, show value, and then drive adoption across the business. ofi, for example, began with us in 2019 and scaled globally once early experiments paid off.
Jason: You previewed an exciting new capability at IFT 2025 in Chicago. What can you share today?
Som: We gave a high‑level preview at IFT. Since joining GlobalData, we can combine Ai Palette’s predictive and generative capabilities with GlobalData’s proprietary retail, foodservice, and market sizing data. That lets us go deeper and more granular than what is possible with public data alone, including packaging‑relevant insights. We are also moving toward multimodal, agentic AI that takes on more of the heavy lifting. The team is hardening the product to GlobalData’s quality standards, with broader availability expected this fall.
Jason: What is next for AI in CPG over the next few years?
Som: So far we have seen functional reinvention. The next wave is enterprise reinvention, where initiatives cut across insights, R&D, supply chain, and manufacturing. After that comes cross‑industry value‑chain reinvention, where suppliers, customers, and sometimes even competitors work differently together. In my view we are still seeing maybe 10% of the potential. The next five to ten years will unlock the rest.
Thank you to Somsubhra GanChoudhuri for sharing how Ai Palette translates consumer signals into products and packaging that win. Real‑world examples show how pairing predictive insight with fast concepting reduces risk, cuts time to market, and sharpens on‑pack decisions from color and claims to structure and materials. We’d love to hear your thoughts: What trends are shaping your innovation pipeline? How are you using AI to inform your packaging decisions?
Ready to bring the future of packaging to life? Or have a perspective on the future of packaging you’d like to share? Reach out to Bob Jennings, CEO of 3D Color, at bob.jennings@3dcolor.com. Our rapid prototyping and color‑perfect samples can help you test new ideas, and we’re always eager to feature fresh voices in Pack Futures.
