Program Case studyLive, multi-locationHuman approval before publish

AI Search Visibility, Review Response, and Customer Lifecycle Automation

Three smaller AI programs running beside the core stack: answer-engine visibility built on topical authority rather than keyword volume, review responses drafted by AI and approved by a person before they publish, and lifecycle sequences personalized on a customer's real history.

Human gatebefore anything publishesTopicalauthority, not keyword volumeAll locationsmulti-location coverageFirst-partyhistory, not merge fields

Built by Patrick Joseph Panuncillon, AI Director and Applied AI Systems Architect. Designed, built, shipped, and operated by one builder with AI augmentation.

The problem

Three quiet losses, none of which look urgent on any given day.

  • Search is shifting from ten blue links to answers. Visibility now means being the source an answer engine cites, which is a materially different problem from ranking, and the old metrics do not measure it.
  • Review responses at multi-location scale are either obviously templated or they do not happen at all. Both cost trust, and the second one costs it publicly and permanently.
  • Customers who deferred a replacement or simply went quiet are the cheapest revenue in the business and the easiest to forget, because forgetting them never shows up as a line item anywhere.

The programs

Three shipped, one scoped and named as such.

AEOAnswer-engine optimization

Visibility built on topical authority rather than keyword volume. The unit of work is covering a subject completely and credibly enough to become the cited source, which is what answer engines reward and what vanity metrics systematically miss.

ReviewsAI drafts, a person publishes

Each review gets a response drafted in the brand's voice, referencing the specifics of what that customer actually wrote. A person approves before anything is published, across every listing location. Nothing auto-publishes.

LifecycleReplacement and reactivation

CRM follow-up sequences with AI personalization for customers who deferred a replacement or have gone quiet, personalized on the customer's actual service history rather than on a merge field with a first name in it.

ScopedVoice agents, specified not shipped

After-hours call answering and appointment scheduling were scoped and specified, including escalation paths and refusal behavior. It is listed here as scoping work, because that is exactly what it is.

01Review arrivesAny listing location
02AI drafts a responseIn brand voice, on the specifics
03Named person reviewsRequired. No auto-publish path exists
04PublishedOr edited, or rejected. A human decides
The review-response path, which is the one where the human gate is not negotiable.

Why the gate is not optional

A published review response is a public statement by the business.

Drafting a review response with AI is a productivity decision, and a good one. Publishing one without a human reading it first is a liability decision, and a bad one. A response that misreads a complaint, concedes something it should not, or misstates a policy is a permanent public artefact attached to the business name.

So the gate is structural. There is no volume threshold, no confidence score, and no quiet-hours exception that lets a response reach a listing without a person approving it. The productivity gain comes entirely from removing the blank page, not from removing the reviewer.

The same principle, three times

Across all three programs the pattern is identical. AI removes the blank page and does the volume work. A person retains the decision that carries public or commercial consequence. That boundary is where the value is, and moving it to save a few minutes is how organizations acquire incidents.

What it demonstrates

The skills behind the programs.

Answer-engine optimizationTopical authority methodologyMulti-location reputation managementHuman-in-the-loop publishing workflowsCRM lifecycle automationAI personalization on first-party historyRequirements and scoping for voice agentsEscalation and refusal designMeasuring conversion outcomes over vanity metrics

Outcome

Live, with the reviewer kept in the loop on purpose.

The search, review, and lifecycle programs run alongside the core stack. In each one, AI does the volume work and a person keeps the decision that carries public consequence. The voice-agent work is specified and is described here as scoping, not as a shipped system.

Confidentiality and intellectual property

This is an anonymized portfolio write-up. It names no client or employer and discloses no source code, prompts, credentials, endpoints, schemas, customer data, or exact commercial terms. Third-party vendors in the client's operational stack are described by category rather than by name. Figures are rounded or banded.

Intellectual property in every system described here belongs to the organization that commissioned it. Nothing on this page is a claim of ownership, a transfer of rights, or an offer to license, sell, or reproduce any system. This page records authorship and build credit only, and it is published in a personal capacity, not on behalf of any employer or client. See Authorship and verification for the full disclosure standard.

If any part of this page is inaccurate, or a rights holder considers it to exceed that standard, contact consultwithpatrick@gmail.com and it will be corrected or removed promptly.