“Helpful, Not Creepy: Personalization That Earns Trust,” Says Kevin McCurdy, Global CPG Partner Lead, AWS

March 11, 2026

Kevin McCurdy, Global Partner Lead, Consumer Goods, AWS, shows how Gen AI, trusted data, and risk-based guardrails turn experiments into repeatable CPG value. He highlights AWS and partner capabilities (Amazon Bedrock, SageMaker, secure integrations) with real wins: demand forecasting, planogram automation, and Adastra’s Mark Anthony Group solution that scales assortment optimization and auto-generates seller scripts. He also outlines quick-win assistants, cost controls, and a company-wide AI program with clear budgets, ownership, and accountability across product, employee, and customer use cases.

• What does it take to move from quick wins with Amazon Q to custom, domain-aware agents on Bedrock that scale across the enterprise?
• When is “good enough” data enough to start, and how can AI assistants surface gaps while improving data quality over time?
• Which operating model and risk-based guardrails help leaders control cost and compliance while accelerating adoption?

Watch the interview:

Read the podcast as an interview:

(The interview was shortened and edited using ChatGPT)

Mark Kohout: Hello and welcome to the Adastra podcast. I’m Mark Kohout, and I lead the governance practice at Adastra, a global AI, data, and cloud systems integrator. We’re in Las Vegas at AWS re:Invent 2025. Joining me is Kevin McCurdy, Global CPG Segment Lead for the AWS Partner Network. Kevin connects AWS with strategic software and services partners to drive innovation for consumer packaged goods and brings over 25 years in supply chain, category management, and demand signal management, with leadership roles at E2open and as a co-founder at Orchestra and Mercari. He’s worked with brands like Coca-Cola, General Mills, Kellogg’s, PepsiCo, Unilever, and Kraft Heinz. Kevin, thanks for joining us.

Kevin McCurdy: Great to be here, Mark. Thanks for having me.

Mark Kohout: Let’s dive in. The big topic today is the CPG industry’s journey to Generative AI and, increasingly, Agentic AI. What patterns and transformation opportunities are you seeing? What excites you most about technology and partnerships in CPG?

Kevin McCurdy: We’re seeing Gen AI and AI/ML applied across core CPG domains: supply chain, sales, marketing, and manufacturing. It’s automating redundant tasks and improving data-driven decision-making in areas like demand forecasting and supply chain planning. AI/ML isn’t new. Companies have used it for decades, but adding “generative” changed the conversation. Two years ago, many worried about job impact; now the tone is excitement and “where can I apply this?” People want to automate routine work so they can focus on higher-value activities.

Mark Kohout: The anxiety is giving way to curiosity and adoption.

Kevin McCurdy: Exactly.

Mark Kohout: How does AWS help CPG companies address their most pressing AI challenges? Where do you provide the most value?

Kevin McCurdy: On the platform side, services like Amazon SageMaker and Amazon Bedrock are foundational. Our role is mapping those capabilities to business use cases. In CPG and retail, we frame this around three areas: product (what’s being sold: discovery, recommendations, optimization), employee (how employees work, knowledge and report summarization, and in CPG specifically, planogram automation), and customer (how consumers engage, search, virtual try-on, and customer service via conversational assistants). We help customers identify and prioritize use cases in each area.

Mark Kohout: The governance person in me thinks about master data and its tight link to the value chain.

Kevin McCurdy: Absolutely. Governance matters at two levels: product master data, aligning how manufacturers and retailers define products, and handling PII for hyper-personalized experiences. Regulations differ by region (EMEA, the U.S., and even U.S. states), and we need to respect them. We also want personalization that’s helpful, not “creepy.” Gen Z expects it, but there’s a balance.

Mark Kohout: Could you share a concrete example where AWS solutions, with a partner like Adastra, delivered tangible impact for a global CPG brand?

Kevin McCurdy: The work Adastra did with Mark Anthony Group is a great example. They faced a common challenge: pulling together sales, assortment, and pricing data, often managed in spreadsheets and silos, to decide what to sell, where, and at what price. Adastra first built a strong data foundation, product, sales, and syndicated data, plus security controls. Then they applied AI/ML for forecasting and prediction. Next, they layered in Gen AI: a sales recommender that not only suggests optimal assortments by market and store but also auto-generates tailored sales scripts, in the seller’s tone and tuned to local demographics and the buyer’s profile. The result is automation and productivity gains, plus freeing sellers to focus on higher-value conversations and opportunities.

Mark Kohout: It’s a compelling end-to-end story, from data foundations to individualized, tone-adapted messaging at the edge.

Kevin McCurdy: And it scales. You’re not relying on one or two star sellers; you can elevate the whole salesforce. The pattern applies broadly across CPG, including apparel, footwear, food, beverage, home, health and beauty, and consumer durables, while allowing each company to leverage its unique data.

Mark Kohout: Let’s talk about partners. I saw that four out of five AWS customers use partners to accelerate AI adoption. How do validated partners complement AWS’s builder culture, deliver quick wins without losing ownership, and where do they add the most value?

Kevin McCurdy: First, education: demystifying generative AI, foundation models, and LLMs. Second, data foundations, governance, and security: partners like Adastra establish the plumbing that makes solutions enterprise-ready. Third, rollout and scale: moving from pilots to production. And finally, change management and training, critical for adoption and trust. This is a new way of working, and partners help organizations build confidence and capability.

Mark Kohout: We’ve seen change management move earlier in the plan, evangelizing and training during experimentation so people are ready at go-live. We also help with use case ideation, POCs, data readiness, vendor selection, and monitoring. From your vantage point, what distinguishes the most impactful partnerships in CPG?

Kevin McCurdy: Domain expertise. Partners who understand the customer’s business, food vs. apparel, supply chain vs. manufacturing, earn trust and deliver better outcomes. Deep data understanding is also crucial: much data is common, but each sector has unique attributes like demographics. And robust knowledge of governance, security, and evolving regulations is essential. The best partners bring all three: domain, data, and governance.

Mark Kohout: Despite the newness of Gen AI, the fundamentals still matter: business at the table, solid data management and governance. Are companies generally ready on compliance and regulatory fronts, or is there still a lot to do?

Kevin McCurdy: We’re still early. The most successful efforts start with a business use case, with business and tech working together. Many begin with a POC; the goal is to reach production and then scale. Governance becomes critical at scale. We’re also seeing customers set rules and contracts around model usage. If you’re using Claude or Nova and want to introduce a different model like Haiku, define what that means before adoption. Explainability and model governance are active areas.

Mark Kohout: Some Gen AI models are black boxes, which raises explainability concerns. At Adastra, our north star is “just enough governance,” a balance between sustainability and overhead, with planning for what’s next.

Kevin McCurdy: Agreed.

Mark Kohout: Last question. For CPGs on the fence, fast followers ready to invest, where should they start with AWS?

Kevin McCurdy: Start with a clear use case tied to value: supply chain, sales and marketing, or manufacturing. Avoid “doing Gen AI to do Gen AI.” Ensure the right data is in place; that’s often the gap. Partners can help unify siloed data and build the foundation for an effective POC, then guide the path to production and scale, reusing what you build for subsequent use cases. Once defined, move quickly.

Mark Kohout: So think strategically, act tactically: build on solid foundations, and design the POC with production and organizational change in mind.

Kevin McCurdy: Exactly, and leverage partners. They’ve done this before, bring expertise in data, security, governance, and regulation, and can accelerate outcomes. As you noted, the vast majority of AWS customers work with partners for a reason.

Mark Kohout: Kevin, this was terrific. I asked for gold nuggets and you delivered gold bullion. Thanks for sharing how AWS and partners are helping CPGs transform with AI. And to our audience, thanks for joining us. If your team wants a low-risk first step, reach out to Adastra and let’s build that quick win together. If you found this useful, please like and subscribe. Bye for now from Las Vegas.

Listen to the podcast: