Got AI FOMO? 3 bold but realistic bets your business can try today (www.zdnet.com)

Pandora’s chief digital officer David Walmsley laid out a pragmatic AI playbook for retailers: avoid FOMO, double down on where AI amplifies your strategic advantage, and run three focused bets now—agentic customer service/sales, model-driven product development, and back-end automation. After an eight-month build with Salesforce Agentforce, roughly half of Pandora’s chatbot traffic already uses next‑generation, agentic agents; the goal is to progress from service to emotionally-aware sales conversations that mirror top store associates. Walmsley frames this as a measured learning journey—reducing customer avoidance of agents and capturing buying emotion are the key challenges. Technically, Pandora blends bespoke and off‑the‑shelf probabilistic models for forecasting manufacturing cost/time and design simplification, and uses a review‑synthesis model trained over two million customer reviews by chaining three third‑party language tools into a custom pipeline. On automation, the company is piloting Microsoft’s Copilot Studio and Teams as workflow front-ends for HR and invoicing tasks to raise productivity and reduce low-value work. For the AI/ML community this is a concrete enterprise blueprint: marry agentic conversational capabilities with probabilistic forecasting and model composition, prioritize integration and UX metrics (e.g., avoidance rate), and focus on ROI-driven, domain-specific deployments rather than chasing hype.