Reputation Tower

Monitoring

35 service groups. One engine.

The monitoring surface of the RepTower loop is one engine with two floors: Open — 20 self-serve groups that document facts and run actions — and Core — 15 groups that judge. Every subject moves through Detect → Adjudicate → Remediate → Monitor, and every cell traces back to a source and a basis. Built for teams that need signals they can defend, not dashboards that shout.

Open — 20 groups
Self-serve tools that document facts and run actions. Every export carries Reputation City branding.
Delisting · Wave detection · Court-grade evidence · Monitoring & alerts · Defense triage · AI visibility · SERM · Google Images · Search suggestions · Geo detection · Social detection · Social remediation · Reviews · Right of reply · Closed-channel detection · Entity normalization · Wikipedia & Knowledge Panel
Core — 15 groups
The verdict engine, delivered — credibility, materiality, coordination, and the score a bank will accept.
Media credibility · Reputation score · Materiality · Field verdict · Compliance shield · Portfolio · Dossier · Demonetization · Infrastructure track · Legal track · Generative visibility · Governance & attestation · Early warning · Deepfake detection · Client portal

Where you start — the flagship views

Seven entry points into the engine; the rest of the catalog sits one click deeper.

Overview
Portfolio reputation in one view: risk levels per subject, active narratives, AI-visibility status and the defense queue.
Monitoring
New mentions, reprint detection and sentiment across the tracked portfolio — the recurring-alert view your team checks daily.
Reputation Score
A measurable, repeatable risk score per subject and geography. Clients see final scores, bands and verdict labels — the methodology stays with the expert body.
AI Visibility
What LLMs answer about a subject, which sources they cite, and what search autocomplete suggests — the GEO axis of reputation.
Coordination
Organic story or seeded campaign? Wave detection with observable signals: duplicate text, fresh accounts, synchrony, shared infrastructure. A campaign verdict is withheld while independent reputable voices corroborate the story.
Defense Triage
Every publication placed on the falsity × public-interest quadrant, with routing, evidence grades and a full audit trail.
Dossier
Biographical dossier readiness: timeline coverage, fact grades A–D, and seven output formats. A fact without a source does not exist.

The field verdict

Signal, not sentence.

The Coordination engine doesn’t answer “clean or dirty.” It classifies the whole information field into one of four verdicts — and this distinction is the moat. Corroborated negativity is never suppressed; keeping it is what makes the rest of the number trustworthy.

Coordinated attack
A look-alike network drives most of the mentions and nothing independent stands behind the story. For a bank: campaign, not substance.
Corroborated adverse
The negative core is backed by independent reputable sources — courts, regulators, tier-1 outlets. We never suppress it.
Mixed field
A real corroborated core alongside a paid-for or low-credibility layer. Enhanced due diligence: separate the two, don't average them.
Clean / organic
No network signals; nothing beyond organic, uncorroborated noise. The wording stays “No signals found in public scope”.

Clients see the class, the band and a plain-language rationale — never the numeric thresholds that produced them.

How we report

Fixed wording, on purpose — so a signal is never mistaken for a verdict.

  • Confidence is always High, Medium or Low
  • “No signals found in public scope” — never “cleared”
  • Signal ≠ verdict. Mention ≠ guilt.
  • Coordination signs ≠ conclusion — expert review decides
  • Internal weights and thresholds never leave the server

Honest about scope

Monitoring supports decisions; it does not make them for you.

  • Continuous monitoring runs as a pilot — scope agreed per portfolio
  • Public sources only — no private-data access
  • Demo walkthroughs use synthetic data and fictional subjects
  • Outputs support human review — not legal or compliance verdicts
  • Scoring methodology is designed for independent assurance — “designed to be independent”, not “is independent”, until that attestation is complete

Questions people ask before they write to us

What do the seven monitors cover?

Overview (portfolio risk in one view), Monitoring (new mentions, reprints, sentiment), Reputation Score (a repeatable score per subject and geography), AI Visibility (what LLMs answer and cite), Coordination (wave detection with observable signals), Defense Triage (every publication placed on the falsity × public-interest quadrant) and Dossier (biographical dossier readiness).

What are the four field verdicts?

Coordinated attack: a look-alike network drives the mentions and nothing independent stands behind the story. Corroborated adverse: the negative core is backed by courts, regulators or tier-1 outlets, and we never suppress it. Mixed field: a real corroborated core alongside a paid-for or low-credibility layer. Clean / organic: no network signals, nothing beyond uncorroborated noise.

How is confidence reported?

Always as High, Medium or Low. A screening result reads “No signals found in public scope”, never “cleared”. A signal is not a verdict, a mention is not guilt, and coordination signs are not a conclusion — expert review decides. Internal weights and thresholds never leave the server.

Does AI Visibility show what ChatGPT says about a subject?

Yes. The AI Visibility monitor tracks what large language models answer about a subject, which sources they cite, and what search autocomplete suggests — the generative-engine axis of reputation, watched continuously rather than checked once.

Can I run monitoring myself, or is it a managed service?

Both, depending on the group. Open groups — delisting, wave detection, court-grade evidence, monitoring and alerts, AI visibility and others — are self-serve in Mission Control. Core groups — media credibility, reputation score, field verdict, compliance shield, dossier and others — carry a human layer. Continuous portfolio monitoring runs as a pilot, scoped per portfolio.

Are the monitor walkthroughs real cases?

No. Every demo published on this site runs on synthetic data with fictional subjects. Client cases are confidential and are never used as demos.

Put your portfolio on the monitors

Pilots run by invitation: we scope the subjects, stand up the surface, and review the first signals together.

Request a monitoring pilot

Confidential. No obligation.