Which Tariff Applies While the Customer Waits on the Line?

What it does

A regulated tariff question reaches the Support Agent while the customer waits, and the rule is unclear. The material sits across tariff sheets, regulatory bulletins, and local operating procedures, for example a connection charge that changed after the last rate case. In Power Utilities a stale answer can commit the utility to a price it may not charge, so every claim must trace to an approved source.

The AI gathers the candidate passages, ranks them by authority and freshness, and maps each claim in the answer to a citation. It checks effective dates, regional applicability, and the agent's permissions before the reply appears. It flags two regional guidances that disagree, a superseded tariff clause, or restricted content the agent may not cite, and returns a conflict notice where nothing holds.

The Support Agent judges whether the answer is firm enough to act on and accepts, corrects, or escalates it. The Finance Controller enters only when the answer changes a charge, a compliance position, or an exception outside the standard tariff. Decision authority remains with the accountable owner, not the model, and the system records the accepted answer and any override.

Where the value comes from

Every hour the Support Agent loses to chasing the current tariff is an hour the customer waits and the queue grows. The benefit is a quicker route from a live pricing question to a cited answer that reflects the rule in force today. In Power Utilities that reads directly as `time_to_answer_reduction` across routine billing and connection queries.

To build the business case, the client measures annual question volume, baseline minutes per answer, and the fully loaded hourly cost of the person who responds. Those inputs drive the primary `transaction_based_time_saving` model, which converts the benchmark range into an annual benefit. Adoption drawn from workflow logs, source acceptance, and escalation rates then shows how much of the projected saving becomes real.

The `search_time_reduction` and `review_time_reduction` views measure the same minutes on the same question population the primary path already counts, so they validate the case rather than add to it. The `knowledge_accessibility`, `decision_traceability`, `compliance_support`, and `expert_dependency_reduction` families have no monetization path and no baseline to price, so they remain control and adoption signals, never a line of savings. That separation is what keeps the number auditable from question to booked result.

How work changes

Manual searching declines for the Support Agent. The role spends less time opening repositories, comparing passages, and asking informal experts about the power utilities policy passage. Repeated documentation of the same answer also decreases when citations are stored. Routine evidence checks become faster and more consistent.

Structured review work increases for the Support Agent. The role checks confidence signals, reads cited passages, and marks whether the grounded answer resolved the question. It also flags source gaps that need ownership attention. This creates better feedback for knowledge owners.

The Finance Controller participates only when the answer affects authority, compliance, legal exposure, or a disputed exception. That involvement should happen through clear escalation rules. Routine questions should stay with the primary role. This prevents the secondary role from becoming an equal reviewer for every answer.

Required skills include source evaluation, evidence reading, and disciplined feedback capture. Accountability stays with the human role that acts on the answer. Freed hours are monetizable only when the client can redirect capacity, reduce backlog, or avoid staffing growth. Managers should monitor whether saved time appears in measurable throughput.