Starbucks CEO Brian Niccol first publicly unveiled two AI systems at Salesforce’s annual conference Dreamforce, including an AI barista tool called “Green Dot” that has already been rolled out at scale in stores—and he also revealed that future apps will be able to “predict your order.”
(Background: Zuckerberg built a “CEO-dedicated AI agent,” and Meta’s 78,000-person company has started letting AI agents “socialize” with each other.)
(Additional background: OpenAI has launched “ChatGPT Agent”! Combining Operator and Deep Research—winning at everything from ticket purchasing and food delivery to writing presentations.)
Table of contents
Toggle
The world’s largest coffee chain Starbucks is quietly rolling out a “secret AI barista” system inside its stores. CEO Brian Niccol revealed at Salesforce’s annual tech conference Dreamforce that the company has already launched two AI systems, covering applications from assisting with drink preparation to order scheduling.
Niccol said that the company’s largest-scale AI deployment so far is an internal system called “Green Dot.” The tool’s functionality is straightforward: when store employees encounter a problem—whether it’s a device malfunction or a special preparation method for a particular drink—they just query the system, and the AI can provide answers immediately, replacing the old process of flipping through manuals or asking veteran colleagues for help.
Since June 2025, Green Dot has been piloted and is now in the phase of large-scale rollout. Niccol described it as “a kind of barista assistant,” and he also emphasized that Starbucks has no intention of taking stores down a “fully robotic” path. He pointed out that the brand’s core remains real craftsmanship and real human interaction, and AI’s role is to assist—not replace.
The second system targets an operational pain point that has long plagued Starbucks: order bottlenecks. Niccol admitted that when he took the job, he found that orders were “processed on a first-in, first-out basis,” creating a large number of bottlenecks. In drive-thru, delivery platforms, mobile order pickup, and in-store ordering, orders from four channels come in at the same time, yet they’re handled with the same logic—resulting in very poor efficiency.
The AI order-scheduling system called “Smart Q” is designed to solve this problem. The system can dynamically adjust the drink-making priority order based on the order source and the customer’s waiting situation. The goal is to have takeaway and drive-thru customers receive their drinks within 4 minutes, while mobile order pickup orders are also completed on time.
Customers can also see three real-time status stages on the screen—“received, in preparation, completed”—reducing the uncertainty of waiting.
Niccol positions the Starbucks app as the core ground for future AI integration. The future scenario he described is simple: users just have to say to their phone, “Hey, the usual order—I’ll be there in 10 minutes,” and the system can predict the order contents based on historical behavior, so that when the drinks arrive, they’re already prepared.
In addition to the customer side, Niccol also said that AI is being tested in parallel on back-end business functions such as visual recognition, inventory management, supply chain optimization, and scheduling—but it has not yet been rolled out at scale.
This series of AI deployments is happening at a key moment in Starbucks’ major strategic transformation. After Niccol took over as CEO last year, he clearly rejected an efficiency-first approach focused on reducing human interaction and expanding purely action-based pickup stores, instead reaffirming the “third place” concept emphasized during the era of legendary founder Howard Schultz—making Starbucks a place people want to stay, beyond home and the office.
In October last year, the company announced a $1 billion reorganization and launched the “Starting 5” plan, selecting five stores as innovation pilot sites to validate before rolling it out nationwide. In the AI era, what kind of results Starbucks will deliver—we can keep watching.
Related Articles
Gate founder Dr. Han: Reshaping the platform’s capability boundaries with an AI ecosystem and multi-asset strategy
A Karpathy-inspired CLAUDE.md Breaks 15K Stars: How a Markdown File Can Tame an AI’s Bad Habits of Writing Code
Bitcoin miners face tough 2028 halving profits, as mining companies deploy AI infrastructure early to address the energy crisis
V God shares: How I built a fully local, private, and self-controlled AI working environment
V God shares: How I build a fully local, private, and self-controlled AI working environment
Collably Network Joins Origins Network to Accelerate Scalable, AI-Driven Web3 Infrastructure