ARTPARK @ IISc | Program
Zetesis Lab: Decision Engineering for Physical Systems
We build verifiable, data-efficient AI that helps teams make high-stakes decisions when reality refuses to cooperate—factories, supply chains, and scientific instrumentation.
Deployed and validated in real systems: Industrial deployments and collaborations include Panasonic, Ather Energy, and research engage ments with DRDO and IIT Delhi (where permitted to disclose).
Decision Engineering: AI that stays right when the world changes
Most operational failures are not prediction failures—they are assumption failures. New configura tions, new suppliers, new constraints, new regulations: the world shifts, and brittle models collapse. Zetesis engineers decision systems that remain defensible under uncertainty and change.
Decision Engineering (definition, business-safe)
Decision Engineering means:
• Mechanism-aware: we model cause-effect and structure, not just patterns.
• Ignorance-aware: we separate what is known, what is plausible, and what is still unresolved.
• Invariance-first: we optimize for decisions that remain valid as reality shifts and new informa tion arrives.
• Verifiable-by-design: constraints and compliance logic are explicit; audits are built in where possible.
What We Deliver
Insight
Make the decision structure legible—drivers, constraints, failure modes, and the assump tions your team is implicitly betting on.
Forecast
Project outcomes under uncertainty—not a single number, but the decision-relevant range and the conditions that move it.
Simulation
Run interventions and stress-tests—“what if we change X?” “what breaks first?” “what keeps us compliant?”