The Future of Enterprise Learning: Integrating LLMs, VR, and Agentic AI at Scale

Training 400,000+ delivery drivers employed by Amazon’s Delivery Service Partners annually across diverse geographies, languages, and learning styles presents challenges that traditional methods cannot solve at scale. This session shares real-world insights from developing and deploying enterprise-grade AI and VR training platforms at Amazon—systems that blend large language models, speech and text recognition, immersive virtual reality simulations, and Agentic AI to deliver personalized, accessible learning experiences that benefit our independent Delivery Service Partners.

You'll learn how we:

• Integrated LLMs and Agentic AI to create interactive learning experiences that adapt to individual learner needs while enabling expert trainers to shift from routine tasks to high-value coaching
• Deployed VR-enabled training tools that provide safe, repeatable practice environments for complex operational scenarios—addressing real-world constraints including noise, connectivity limitations, language diversity, and varied technical capabilities, improved the knowledge retention rate by 70%
• Improved training preparation time to readiness by 30% while strengthening competency scores through AI-powered platforms that enable location-independent, self-paced learning
• Applied operational discipline from manufacturing and supply chain leadership to AI development—ensuring reliability, explainability, and progressive disclosure in user interfaces
• Balanced innovation with ethics, designing for augmentation and maintaining dignity throughout digital transformation

This presentation offers practical frameworks for organizations seeking to deploy XR and AI technologies at enterprise scale, including lessons learned on cross-functional collaboration, risk management, compliance monitoring, and change management. Whether you're building training systems, operational support tools, or customer-facing AI applications, you'll gain actionable insights on bridging the gap between cutting-edge technology and real-world implementation.

Key Takeaways:

• Architectural patterns for integrating LLMs, VR, and Agentic AI in enterprise learning platforms
• Strategies for designing inclusive, accessible AI systems that work across diverse user populations and operating conditions
• Frameworks for measuring impact beyond performance metrics—including training effectiveness, learning outcomes, and operational resilience
• Approaches to AI deployment that prioritize human expertise