01 — The headline
From truck-first activation to governed remote intent
The product work was to close the gap between customer-service intent and confirmed meter outcome. We locked a system contract that targets under 24h connects where eligibility and safety rules are satisfied.
02 — Situation
A latency problem hidden inside operations complexity
Move-ins were often scheduled as truck events, creating avoidable delay and customer friction. Meanwhile, AMI readiness, ID consistency, and policy branching were fragmented across teams and tools.
03 — Constraints
Constraint stack for RCD
RCD is only useful if failures are explicit and safe, so requirements had to encode operational truth, not optimistic command paths.
Safety and policy branchingCollections, emergency, move-in, and move-out each required different control logic and escalation behavior.
Device and network varianceElectric and gas meter classes with uneven connectivity meant command reliability could not be assumed.
Cross-system identity alignmentRCD cannot work when CSS and AMI inventories disagree on premise or meter identity.
Audit and customer truthEach attempt needed evidence-quality outcomes that operators could explain and defend.
04 — Key decisions
RCD framing decisions
We narrowed scope to decisions that preserved trust at scale, even when remote action failed.
Define explicit failure semantics
The callStandardized retry, timeout, and exception outcomes instead of silent reattempt loops.
The tradeoffMore upfront process detail to reduce downstream operator confusion and duplicate actions.
Treat eligibility as a first-class gate
The callBuilt scenario-based eligibility matrices before command orchestration details.
The tradeoffSome implementation teams waited on policy lock before prototyping edge features.
Sequence by operational risk
The callPrioritized high-volume, lower-risk scenarios first and codified truck-roll fallbacks early.
The tradeoffDeferred broad novelty in favor of credible phased adoption.
05 — What shipped
Discovery artifacts that engineering can execute
The shipped output in this phase is a reconciled requirements model: initiator contracts, eligibility rules, ack semantics, and governed exception routing across AMI and customer-service systems.
Callout: discovery here is not abstract research; it is the implementation-grade map for safe remote service at utility scale.
06 — Results
Readiness outcomes to date
Before broad launch, the measurable result is reduced ambiguity: teams now operate from one requirements baseline with explicit non-remote paths.
07 — For the next PM
What this teaches
When hardware capability is uneven, product clarity around intent, evidence, and fallback is what unlocks real adoption.
Design for verified outcomesOperational trust depends on confirmed state transitions, not command volume.
Policy language must become product languageTranslate scenario ambiguity into explicit eligibility and operator paths before scaling delivery.
Fallbacks are part of the featureTreat non-remote routes as first-class to avoid brittle adoption during edge conditions.