Barry Callebaut launches AI innovation center for chocolate in Singapore - here's what product teams can do with it
Barry Callebaut has opened a Global Innovation Center (GIC) at Geneo in Singapore Science Park, pairing the "world's first" AI Center of Excellence for chocolate with a new Cacao Coatings Center of Excellence. The goal is simple: move from idea to scalable product faster, with data guiding every decision from flavor mapping to process fit.
- AI in the loop: Predictive flavor development, formulation guidance, and personalized customer solutions.
- Speed to pilot: State-of-the-art pilot lab at the Senoko factory for rapid prototyping, small-batch runs, and process testing.
- Category focus: Strong emphasis on coatings and fillings for applications like ice cream, pastries, and snacks.
- Regional advantage: Singapore's talent pool, infrastructure, and proximity to AMEA growth markets.
Inside the Singapore GIC
The AI Center of Excellence is purpose-built for chocolate and cocoa R&D. Expect data-driven concepting, faster iteration loops, and tighter links between sensory feedback, formulation choices, and predicted consumer appeal.
The Cacao Coatings CoE is Barry Callebaut's first hub fully dedicated to coatings and fillings (compounds). It focuses on breakthrough formulations and application expertise, with a clear path to scale across brands and markets.
On-site, the Callebaut Chocolate Academy Singapore teams with a regional R&D facility to turn customer briefs and trends into testable concepts. A new pilot lab at the nearby Senoko chocolate factory connects the dots from benchtop to line trials.
Why this matters for product development
Volatile cocoa prices and softer demand are forcing smarter bets and faster cycles. The GIC gives teams a place to co-create with data, quickly validate product-process fit, and de-risk scale-up before committing to full runs.
"AI is redefining how we imagine, design, and deliver chocolate and cocoa experiences," says Amr Arafa, chief digital officer. For developers, that means less guesswork and clearer evidence for go/no-go decisions.
How to engage (and get value fast)
- Bring structured briefs: Target consumer, sensory profile, claims, cost-in-use, process constraints, and line capabilities.
- Share usable data: Past launch performance, panel results, stability data, and key retailers/channels to feed AI models.
- Prototype with purpose: Use the Senoko pilot lab to lock specs, validate machinability, and stress-test shelf life before scale.
- Set hard KPIs: Define success on predicted liking, formulation cost, time-to-first-PO, and yield at trial.
- Plan compliance early: Align on regulatory, allergen, and label claims to avoid reformulation late in the cycle.
Team and capabilities on site
The center will host 30+ specialists: AI engineers, food scientists, chefs, application technologists, process technicians, and customer experience leads. Practically, that means cross-functional sprints-from concept to pilot-can happen in one place.
Coatings and fillings: where the near-term wins are
Coatings and fillings drive sensory impact and cost control. Expect focus on snap, melt, viscosity, bloom resistance, bake stability, and enrobing/compound performance-optimized for lines in confectionery, bakery, and ice cream.
Broader moves you should know about
- Factory upgrades: €250M for Wieze (Belgium) and €125M for Halle to modernize lines and improve safety-key for reliability at scale.
- AI partnerships: Collaboration with Chile-based NotCo to speed recipe development using AI.
- Cocoa-free options: Long-term partnership with Planet A Foods to scale ChoViva, a sunflower- and oat-based chocolate alternative.
Market context and product strategy
Industry data shows value up, volume down-price-driven revenue is masking weaker consumption. Barry Callebaut's "Top Chocolate Confectionery Trends 2026 & Beyond" highlights a shift to smaller, everyday treat moments-what the company calls "Minorstone Confectionery." That suggests portionable formats, impulse-friendly pack sizes, and affordable premium cues will matter more in briefs.
If you're planning a project
- Best-fit projects: New coatings/fillings, reformulations to hit cost targets, flavor map extensions, line-fit optimization.
- Pilot timeline: Use the Senoko lab for quick sprints-prototype, line trial, iterate-before locking final specs.
- Data handshake: Clarify data rights and feedback loops up front to keep the AI models learning from your trials.
How to start
- Connect with Barry Callebaut's regional team or the Chocolate Academy/GIC network to scope a brief.
- Prepare two tracks: a coatings/fillings workstream and a packaging/channel workstream to match "Minorstone" occasions.
- Set a pilot gate: proceed to scale only if predicted liking, yield, and cost-in-use hit targets.
Upskilling your team on AI for product work
If your roadmap depends on AI-assisted formulation and faster decision cycles, consider a lightweight training path for your R&D and innovation leads. Here's a curated list of AI learning options by role: AI courses by job.
Barry Callebaut's bet is clear: integrate AI, tighten the loop between concept and plant, and give product teams a faster path to market-especially in coatings and fillings where speed, cost, and sensory impact meet.
Your membership also unlocks:
