• Architect agent frameworks capable of reasoning across multiple steps and data sources • Collaborate directly with enterprise clients to understand reasoning needs, workflows, and tool integration requirements • Design planning and decision-making loops that integrate retrieval, tool use, and external APIs • Implement strategies for multi-hop reasoning, synthesis, and constraint-based problem-solving • Ensure retrieval outputs are structured for optimal reasoning use • Evaluate agent performance across reasoning benchmarks, accuracy metrics, and task completion rates • Guide engineering teams on agent orchestration platforms and best practices • Maintain playbooks and reusable reasoning templates for client-specific workflows