Gary A. Stafford·Aug 21Multi-Agent Orchestration, MCP, and Human Approval: Mars Mission Control with the OpenAI Agents SDKAn interactive crisis simulation demonstrating the OpenAI Agents SDK’s core features, including multi-agent orchestration, MCP, structured…
Gary A. Stafford·Aug 9The Right Model and Reasoning Level Depend on the WorkloadHow workload and reasoning level shape the cost of a correct answer.
Gary A. Stafford·Aug 4Valuemaxxing Reasoning Tokens: The Cost and Correctness of Frontier ModelsA practical benchmark of how model choice and reasoning effort affect token consumption, cost, and answer quality on Amazon Bedrock Mantle.
Gary A. Stafford·Jul 24Building a Web Search-Powered Agent with Amazon Bedrock AgentCoreGive AgentCore Harness secure, up-to-date web search capabilities with Amazon Bedrock AgentCore’s Web Search tool
Gary A. Stafford·Jul 19Capability Elicitation and Emergent Behavior in LLM SystemsA Practical Framework for Emergent Behavior Across Models, Inference, and Systems
Gary A. Stafford·Jul 16Automating Model Training and Delivery from Amazon SageMaker AI to Amazon EKSAn end-to-end walkthrough of model tuning, evaluation, explainability, registration, and approval-driven Kubernetes deployment.A response icon1A response icon1
