Q&A with Vishal Singhvi at The Heineken Group
Heineken is driving digital transformation in the brewing industry through its Global GenAI Lab in Singapore, led by Vishal Singhvi. The Lab develops scalable generative and agentic AI solutions to boost growth, productivity, and customer engagement across 70 markets and all business functions. Multidisciplinary teams accelerate innovation in areas like autonomous agents, automated marketing, advanced financial tools, and intelligent customer support, combining startup agility with enterprise scale. Vishal, the Global Head of GenAI & Agentic Innovation, brings deep expertise from leadership roles at Microsoft and Mastercard, where he translated AI ambition into real-world impact. Vishal will be a guest speaker at MRC’s Melbourne event on June 23, 2026, covering agentic commerce trends, open standards for retailers, strategies for brands in an agent-mediated world, and real-world examples of how agentic commerce can be a competitive advantage.
BMW: Could you give us a high-level overview of The Heineken Group, and describe your role within the company?
VS: HEINEKEN is one of the world’s leading brewers, with a global footprint, strong international and local brands, and an end-to-end value chain covering brewing, supply chain, sales, marketing, customer engagement and sustainability.
I lead HEINEKEN’s global work on generative AI and agentic AI from our GenAI Lab in Singapore. My focus is simple: move AI from pilots to business impact. That means identifying high-value use cases, building reusable capabilities, supporting adoption across markets, and ensuring AI is deployed responsibly and securely.
I also bring an external commerce and risk lens through my role with the Merchant Risk Council in APAC, where I work with leaders across payments, fraud, digital commerce, and agentic commerce.
BMW: How is Heineken leveraging artificial intelligence to enhance its global operations?
VS: We use AI as a practical business tool, not as a technology experiment. The test is always: does it improve growth, productivity or future-fit capability?
Across commerce, AI helps teams make better decisions on sales execution, customer opportunities, promotions, investment allocation and customer engagement. In operations, it supports predictive maintenance, quality control, line optimisation, logistics, and resource efficiency. In finance and procurement, it improves forecasting, spend visibility, data quality and scenario planning.
Generative AI helps colleagues find information faster, summarise documents, draft content and access internal knowledge. Agentic AI is the next step: systems that can monitor signals, recommend actions and support controlled workflows.
The principle is human augmentation. AI should remove repetitive work, improve decision quality and give colleagues more time for customers, creativity, quality, and innovation.
BMW: What advantages has Heineken seen from integrating real-time data into its operational systems?
VS: Real-time data changes the operating model. It helps teams move from monthly reporting to faster sensing and response.
In breweries, live signals can help detect issues earlier, reduce downtime, and improve performance. In logistics, they support better planning and service levels. In commerce, they help teams respond faster to demand shifts, promotion performance, and execution gaps.
For AI, real-time governed data is foundational. Without consistent, connected data, AI does not scale well. With strong data foundations, AI can move beyond explaining what happened to recommending what teams should do next.
BMW: How is the deployment of AI supporting Heineken’s pursuit of its “Brew a Better World” sustainability goals?
VS: AI can support sustainability, but we need to be disciplined and transparent.
There are practical use cases. In breweries, AI can help improve energy and water management, detect deviations earlier, optimise cleaning cycles, and reduce losses. In logistics, it can reduce inefficiencies. In packaging and circularity, it can improve the visibility of returnable flows and identify leakage points. In reporting, it can improve data quality, traceability, and anomaly detection.
At the same time, generative AI has its own environmental footprint. It consumes computing, energy and water through the infrastructure behind it. So, our approach is to use AI where the value is clear, avoid wasteful usage, right-size the technology to the task, and improve measurement over time.
AI should help us make better decisions with less waste. It should not become technology without accountability.
BMW: What are the most significant changes you’ve observed in the agentic commerce landscape since 2025?
VS: The biggest change is that agentic commerce has moved from experimentation to production.
In 2025, most conversations were about AI shopping assistants. In 2026, the infrastructure is becoming real: open commerce protocols, agent-ready checkout, tokenised payments, cryptographic consent, agent identity and real-time fraud signals.
The purchase journey is being rebuilt. The consumer can now delegate discovery, comparison, selection, checkout, and post-purchase tasks to an AI agent. That means there may be no search tab, no traditional checkout page, and no manual input.
For merchants, the strategic question is simple: who controls the transaction rail? If the platform owns the mandate, it can influence the customer relationship beyond the transaction. If the merchant owns the mandate and the post-purchase experience, they keep the relationship, the data, and the trust layer.
BMW: How do open standards—like UCP and various payment protocols—work, and what are the key benefits they offer to retailers?
VS: Open standards create a common language between agents, merchants, platforms, and payment networks.
They help merchants avoid building one-off integrations for every AI surface. A merchant can expose structured product, inventory, delivery, checkout and policy data once, and make it usable by multiple agents.
The benefit is reach with control. Reach, because agents can discover and transact with the merchant more easily. Control, because the merchant can remain seller of record, keep transaction visibility, preserve payment mandate records, and own the post-purchase journey.
The payment protocols add the trust layer. They can define what the consumer authorised, what limits apply, which cart was approved, and how payment tokens are used without exposing raw credentials to the agent.
The goal is not just faster checkout. The goal is safer delegated commerce.
BMW: Could you share some real-world examples of agentic commerce in action, such as payment automation and personalised shopping agents?
VS: A practical example is Australia’s first authenticated agentic transaction. A consumer asked an AI agent to buy movie tickets using a bank card. The transaction used tokenised payment credentials, the agent did not see the raw card number, and the merchant still received payment and transaction visibility.
A second example is a large retailer using open protocols. A consumer could ask: “Find the best 65-inch OLED TV under $3,000, deliver it to Melbourne by Friday, and use my card.” Without open standards, the agent may scrape pages, hit bot controls, and fail at checkout. With structured commerce data and payment mandates, the agent can read inventory, check delivery, confirm eligibility, generate an authorised cart and complete payment safely.
The pattern is consistent: the agent becomes a new authorised channel, but fraud controls, token controls, chargeback processes and merchant visibility still matter.
That is the right model: delegation with guardrails.
BMW: You’re speaking at the upcoming MRC Melbourne Event on June 23 about the agentic commerce landscape. Without revealing too much, could you briefly address some of the core topics your session will cover?
VS: The session is called “Open Protocol, Owned Experience: Winning in Agentic Commerce.”
I will focus on three questions. First, what changed: agentic commerce is no longer a future trend. It is moving into live transactions. Second, what is at stake: merchants risk becoming invisible if agents cannot read their data, or becoming only fulfilment if platforms own the customer mandate. Third, what to do now: get discovered, own the mandate and own the relationship.
For payment and fraud teams, the role expands. They need to verify agent identity, monitor agent-initiated transactions in real time, manage liability, and protect the post-purchase experience around disputes, loyalty, returns and support.
The core message is simple: open protocols can create reach, but merchants still need to own the customer experience.
BMW: How can forward-thinking brands turn agentic commerce into a competitive advantage, rather than seeing it as a potential threat?
VS: Brands should treat agentic commerce as a new operating channel, not a chatbot feature.
There are three moves. First, get discovered. Product data is the new storefront. If agents cannot read your catalogue, they will not recommend it to you. Structured, enriched, machine-readable data becomes as important as search optimisation.
Second, own the mandate. Brands and retailers should avoid giving away the authorisation layer too early. The mandate, payment token lifecycle and transaction record are strategic assets.
Third, own the relationship. Agents can assist with discovery and purchase, but they cannot replace post-purchase trust. Returns, loyalty, customer support, personalisation and service recovery remain in the merchant’s moat.
The winners will not be the brands that chase every agent platform. The winners will be the brands that are easy for agents to understand, safe for agents to transact with, and strong enough to keep the customer relationship after the purchase.
BMW: You’ve just joined MRC’s APAC Advisory Board, offering valuable insights from your region on payments, fraud trends, and industry innovation. In what ways do you see yourself supporting and guiding the MRC over the next year?
VS: MRC is valuable because it brings merchants, payment providers, technology companies and fraud leaders together to solve shared problems.
APAC is one of the most diverse commerce regions in the world. We have cards, wallets, real-time payments, super-apps, cross-border commerce, mobile-first consumers and very different regulatory environments. That makes the region a strong signal for what is coming next.
I hope to support MRC in four ways. First, bring an APAC perspective into global discussions on payments, fraud, and agentic commerce. Second, help connect the AI and payments communities. Third, contribute thought leadership on AI-enabled fraud, synthetic identities, deepfakes, automated social engineering, first-party misuse, dispute abuse and agent-led transaction risk. Fourth, help members move from awareness to readiness through practical frameworks, education, and peer learning.
Commerce will become more automated, intelligent, and real-time. The challenge is to make it more trusted at the same time. That is where MRC can play a meaningful role.
***
Author: Bernadette Walsh, Marketing Partnerships Manager, Melbourne, Payments Consulting Network
Bernadette’s diverse experience has given her the opportunity to observe various approaches to business and problem-solving, helping her develop a unique perspective on building relationships and addressing challenges. Her resourcefulness and willingness to think outside the box have been invaluable in finding creative solutions to complex problems.
***
Payments Consulting Network is a media partner of Merchant Risk Council, in Melbourne 23 June 2026. Be Part of the Future of Payments and Fraud Prevention – sign here to register.
Other upcoming MRC conferences:
***
If you found this article helpful and would like to read similar articles, please subscribe to our newsletter.
