CLIENT
INDUSTRY
Consumer markets
FEATURING
AI
Product and platform engineering
500+
processes and data sources mapped, spanning the Global Sustainability Office
17
agentic loops designed across core process areas
43
AI capabilities identified and prioritized for implementation
PwC helped design an AI-native operating model so PepsiCo could scale sustainability faster and more efficiently
SITUATION
Sustainability meets the AI moment
PepsiCo’s Global Sustainability Office (GSO) drives pep+ across three interconnected pillars focused on driving positive impact and building business resilience. Behind these pillars lies a complex operational reality. More than 500 distinct processes and data sources span procurement, research and development, marketing, manufacturing, finance and external reporting. As PepsiCo accelerates its agentic strategy, sustainability workflows should transform to keep pace.
The status quo was challenging. Forecasting and planning could take weeks to complete, limited by data availability, modeling complexity, and the lack of enterprise tools available industrywide in the sustainability space. At the same time, stakeholders increasingly expected timely, transparent sustainability information, and AI was changing how that information could be analyzed and used.
PepsiCo’s leaders knew the GSO needed to evolve into an AI-native operation that could help the business adapt faster, make more-informed decisions, and keep pace with rising stakeholder expectations. PepsiCo turned to the team at PwC, who understood pep+ and how AI could help evolve it for the future. With the help of PwC, PepsiCo defined a vision, built core agentic AI capabilities, and created a comprehensive suite of resources to help drive enduring transformation.
SOLUTION
Designing a blueprint for an AI future
PwC and PepsiCo structured the transformation around three priorities: mapping the GSO’s operating processes; designing AI-enabled workflows; and creating the tools, governance, and roadmap to scale adoption across the function. We started by creating a visualization of 500 nodes representing the processes, data sources and decision points that make pep+ run. The map revealed hidden dependencies and redundancies that frequently created bottlenecks.
What started as linear workflows transitioned to 17 AI-forward, agentic-driven loops. The old model executed steps in sequence and then stopped, whereas agentic loops operate continuously. The loops can detect changes in data and context, synthesize implications using AI reasoning and then take action, either by flagging it for human review or triggering the next step. They are supported by eight easily understandable agentic archetypes spanning 40-plus capabilities, handling functions like data validation, calculation, and orchestration. Because archetypes are reusable across loops, efficiency grows as the portfolio scales. This can simultaneously further adoption and advance the function as a whole. The result is a more adaptive sustainability operating model that can respond more quickly to changing business conditions, regulations, and stakeholder needs.
This shift toward an intelligent infrastructure can create a foundation for what comes next.
Analyzing the data, validating the model
PepsiCo’s leaders knew where they wanted to deploy AI, and the analysis confirmed their instincts. Two domains stood out: where the value case was clearest and where the design lessons would be transferrable.
- PepsiCo pep+ glidepath and delivery
Driving progress toward sustainability goals requires regular forecasting to reflect evolving input data and market conditions. With the help of PwC, PepsiCo built an interactive blueprint for automated scenario analysis that can recalculate glidepath projections in real time as variables change. A multi-week analysis can now be triggered on demand. - Narrative and messaging
Sustainability analyses across the function draw from similar data, yet teams routinely start from different points. We designed an AI-enabled data management layer with a verified information repository. An agent will assess data, check for completeness and detect anomalies, allowing team members to surface more consistent insights faster.
Making it stick
Engineering an AI-native operation takes more than use cases. AI is constantly changing, and we need to build in an agile way so that our designs continue to evolve as technology advances. PwC created a framework to sustain the transformation, pairing a clear vision with detailed plans.
It’s anchored around a four-phase implementation roadmap and AI builder playbook, a custom user manual for building internal capabilities. The framework is designed to equip the GSO team to identify opportunities, expand on agentic loops, and apply industry leading practices for prompting and governance. We also designed a governance framework with clear protocols for performance monitoring, investment decisions, and human oversight—all aligned with PepsiCo’s broader responsible AI policies. These human-centered capabilities will help empower PepsiCo to continue to help drive their transformation.
Get more on this topic
RESULTS
Positioned to lead with precision, scale, and speed
The clearest signal that the new operating model is working? The people. Scenario analyses that once took weeks now run in minutes, freeing PepsiCo’s professionals to focus on higher-value activities, enhancing the value and impact of the function. When a packaging input cost shifts or a senior leader poses a what-if question on a Friday afternoon, the GSO can help find the answer. Teams can make smarter, real-time decisions as situations evolve, freeing up capacity for more-complex planning work.
But speed alone means little without the trust that’s engendered by true, clean data that underpins analysis and decision-making. Automated anomaly detection can catch inconsistencies earlier, improving the integrity of the data. While humans remain firmly in control of the final product, AI-assisted drafting is starting to accelerate sustainability analyses. The model is becoming a force multiplier for pep+ as team members are empowered by both AI capabilities and the upskilling and hands-on practice that allows them to continue to transform and grow. The GSO team can identify opportunities, mock up ideas, and collaborate more impactfully with their internal and external stakeholders.
Compounding returns
One of the most important outcomes is a repeatable approach taking shape across PepsiCo. The GSO is becoming a proof point, building a referenceable model enterprise-wide. AI enables pep+ to operationalize sustainability in ways that help reduce inconsistencies and support growth, while maintaining focus that the positive business and sustainability outcomes enabled by AI outweigh its associated energy and water footprint. And, as more agentic capabilities go live, time and capacity savings compound, unlocking new opportunities with each cycle. This model shows how sustainability and profitability can reinforce each other, creating resilience, efficiency, and long-term value.
This is one of the first efforts to transform corporate sustainability into an AI-native operation. It goes beyond PepsiCo’s deep grounding in sustainability. It demonstrates how AI can help embed sustainability into decision-making at scale. And it’s a glimpse of what’s possible when expertise, technology, and ambition move together.
Get more on this topic
The uncomfortable truth about AI and workforce transformation
“PwC understands our business and pep+, and that added value from day one. This model shows how sustainability and growth can reinforce one another at scale.”
“PepsiCo Positive is core to our strategy. AI can help accelerate pep+ and become a force multiplier to drive progress. PwC helped us build the model to stand up key AI capabilities and the roadmap to leverage AI for positive change in a big way.”
WHAT COULD AI UNLOCK FOR YOUR COMPANY?
Rethink sustainability for greater speed, scale, and business value.
CASE STUDIES. REAL IMPACT.
Expertise. Technology. Results. Powered by collaboration. Explore our case studies to see what’s possible.
