168 to make predictions of future trends,” she explains. “However, we have witnessed that organisations are still in need of more realtime insights. This is where agentic AI can make a difference.” Solutions for change Microsoft provides the technology foundation, with products such as Microsoft Azure, Dynamics 365 and Fabric. Meanwhile, the Microsoft partner ecosystem brings the RGM domain expertise and implementation experience. Through this collaboration, they help CPG firms unify data, automate insights and orchestrate decisions across pricing, promotion and trade investment. For example, The Xtel AI Platform harnesses the power of AI to help CPGs manage complex data, address intricate questions, develop sophisticated commercial strategies and optimise enterprise planning and execution in real-time. According to Piet Surmont, global head of strategy and analytics at at Xtel, AI agent personas in RGM act as a “team of consultants”. “The user interacts with a multi-agent system, and each agent specialises in tasks, such as impact analysis of price changes, and collaborates to provide real-time, actionable answers,” he explains. “Max.AI unifies sales by providing a strong data foundation, explainable AI, deep business context and a powerful planning engine. It breaks down negotiations into sub-goals such as price optimisation and margin security, then delivers actionable strategies.” Powering innovation in RGM As the RGM journey continues to evolve, Marco Casalaina, vice president of core AI at Microsoft, offers his perspective on how generative and agentic AI are reshaping the landscape of machine learning, data science and data management automation. Casalaina sees the future of RGM unfolding at the intersection of generative AI and classical machine learning. This convergence is transforming predictive analytics, with new generative models for time series forecasting and anomaly detection opening doors to deeper business insights. He emphasises that data science agents, capable of automating feature engineering and model setup, are streamlining the creation of predictive models. For consumer goods companies, this innovation promises to scale analytics efforts and unlock new efficiencies. At the heart of this transformation is the role of conversational data agents. “Data agents can allow you to chat with your data, which is useful by itself, but the real value of them is to endow downstream agents with data,” he explains. “It lets you encapsulate the data querying logic so your other agents can focus on the jobs they need to do.” The roadmap for adoption When it comes to agentic AI adoption, Praneet Aneja, chief of staff and product strategy at Fractal subsidiary Asper.AI, says the process in RGM is a journey which will evolve in three stages. The first, and current, stage is RGM as a copilot. By building RGM domain agents that can generate insights on the data and AI model outputs provided, RGM experts can become more productive and handle a bigger mandate, while driving more value for markets where there is no dedicated expert. The second stage focuses on collaborative intelligence. Training the agents to understand the enterprise functions such as marketing, sales and supply chain will help to drive better alignment between cross-functional teams, increasing speed of trust and alignment. This will help to unlock more productivity and improve RGM mastery. The third and final stage focuses on autonomous decisions. “This will be the stage where agent design and training will extend to include deep research on external or enterprise information, reasoning models, fine tuning of domain agents, perform modelling and simulation for different scenarios,” says Aneja. “This will start with low-critical and high-volume decisions, with success leading into mid- to high-critical decisions and unlock direct business impact on gross margins, volume and sales.” FEATURE “ Leaders must reimagine RGM as a dynamic, intelligent and agentic capability” DINA ZHOU, MICROSOFT
RkJQdWJsaXNoZXIy NzQ1NTk=