The AI Engineering Stack: What Actually Belongs Around the Model
Saleem Chauhan
The model is only one component of a production AI system. A practical architecture for orchestration, context, tools, policy, evaluation, observability, security and cost.
September 2026
From AI Demo to Production: What Changes When the System Has Consequences
Moving AI from an impressive demo into production changes the engineering problem. A practical framework for evaluation, boundaries, observability, recovery and controlled autonomy.