-
Book Overview & Buying
-
Table Of Contents
30 Agents Every AI Engineer Must Build
By :
This chapter examined two complementary architectures for building agents that are not only capable but also responsible: Ethical Reasoning agent and Explainable agent. We first learned how every agent action is evaluated against explicit ethical constraints before execution through deontic logic and quantitative fairness metrics. The Impossibility Theorem demonstrated that fairness requires deliberate, documented choices about which metric to prioritize, and the HR assistant case study showed how these principles translate into a practical system that prevents demographic bias while maintaining evaluation quality.
The Explainable agent addresses the equally critical challenge of transparency. By recording the agent's reasoning trace, generating audience-adapted explanations using LIME, SHAP and counterfactual analysis, this architecture transforms opaque decision-making into a transparent, auditable process. The medical diagnosis assistant illustrated how these techniques...