Palantir Technologies (PLTR) investor relations material

Keynote themes and industry context

• AI is driving a new era of operational transformation, with sovereignty, security, and domain-specific intelligence as central themes.

• Forward-deployed engineering and shared ontologies are enabling organizations to break down silos and accelerate decision-making.

• The conference highlighted the need for continuous learning loops, human-AI teaming, and the importance of owning data and models.

• Industry leaders emphasized the shift from reactive to proactive, predictive operations across sectors.

• Partnerships between technology providers and enterprises are reshaping supply chains, safety, and customer engagement.

AI-driven supply chain transformation

• NVIDIA and Palantir built a Command Center using a shared ontology, integrating data from suppliers, manufacturers, and operations for real-time visibility and optimization.

• cuOpt and Nemotron models, fine-tuned with proprietary data, enable scenario planning, rapid retraining, and continuous improvement in supply chain decisions.

• Sensitive data and decision logic remain on-premises, ensuring sovereignty and security while enabling rapid scaling and retraining.

• Weekly performance metrics and transparent dashboards are replacing traditional quarterly reviews, driving supplier accountability and ecosystem-wide optimization.

• The approach is being extended to partners, aiming for a globally optimized, collaborative supply chain network.

Human-AI teaming and operational excellence

• Human expertise is central, with AI agents codifying tribal knowledge and supporting analysts in high-value, exception-driven tasks.

• Decision logs and feedback loops are used to benchmark, evaluate, and continuously improve models and workflows.

• The system enables rapid scenario analysis, root-cause investigation, and proactive risk mitigation, reducing manual effort and accelerating response times.

• Security and governance are embedded, allowing organizations to move faster without compromising compliance or control.

• The model supports both deterministic and stochastic intelligence, blending optimization algorithms with LLMs for robust decision support.