BharatRouter

Experiments

Run a golden dataset across prompt variants — see which wins on pass-rate, cost and latency, then promote it.

Pass-rate

Share of rows each variant passed. Winner in accent.

Pick an experiment.

Mean cost

Notional ₹ per row (usage-value, not a bill).

Latency

p50 / p95 response time (ms).

Per-variant

One row per variant. Promote repoints the winning prompt's prod tag. Owner/admin only.

VariantPrompt / modelPassedPass-rateMean ₹Total ₹p50 msp95 msErrors
No run selected.

Notional ₹ is usage-value, not a charge. Experiment traffic rides the loopback self-call so routing, prompt-render and residency pinning apply exactly as production.