Engineering Trust: Building Reliable AI Systems for Business Operations
Analytics Insight, Tuesday, November 3rd, 2026
MiQ India's Rohit Srivastava argues AI reliability comes from the whole decision pipeline, not the model, as systems become agentic.
Rohit Srivastava, VP Engineering at MiQ India, notes that a model can pass testing and still fail in production because of stale data, shifted users, untracked policies or unwatched workflows.
Reliability is therefore a property of the whole system.
As AI moves from recommending to planning, deciding and acting, silent failure becomes unacceptable, so the unit of trust is the full pipeline of data, features, model, policy, action and feedback.
He describes how data quality, observability, governance, human oversight and cost control help businesses build AI systems they can rely on in live operations.