RAG Signal
Platform

40+ Modules. One Objective: Get You Cited.

Our team operates this proprietary system during your engagement — tracking the evidence, technical work, and reporting across ChatGPT, Claude, Perplexity, and Gemini.

Live Dashboard

Built In-House. 40+ Modules.

The team uses this environment to run and report client work. It is not another dashboard for you to manage.

Citation Intelligence — client view
Citation Rate
81%

51 of 63 prompts citing brand

Per-Model Citation
ChatGPT84%
Perplexity86%
Claude78%
Gemini76%
Brand Memory Health
87 /100
Entities Active42
Signal Alerts0
Last Sync2h ago

Actual dashboard — anonymized. Every client sees their own real-time data.

The team’s working environment

We do not sell you a platform. We operate it for you.

The RAG Signal team uses the system we build to collect evidence, set priorities, direct technical work, and report the result.

Evidence graph — RAG Signal client-work environment
Evidence graph

We connect the evidence behind Brand Memory.

Sources, entities, and verifiable claims are organised into one working context for our team to evaluate.

Prompt Discovery — RAG Signal client-work environment
Prompt Discovery

We prioritise the buyer questions worth tracking.

Evidence, candidates, visibility measurement, and the roadmap move through one team-operated cycle.

Citation tracking — RAG Signal client-work environment
Citation tracking

We turn signal movement into a report.

Our team interprets model, prompt, and source movement so you receive the context needed for a decision.

These screens show the environment used during client work. We provide expert-led delivery — not a dashboard you are expected to operate.

Question crawled sources Source A Source B Source C Answer retrieved passage not retrieved
The unit that gets selected is the passage, not the page. Being relevant to the topic is not enough — the source has to contain a section that answers the question on its own. Source C may well be about the subject, but none of its sections answers this question directly.
CI

Citation Intelligence

Real-time tracking of your citation rate across every prompt and model. See exactly where you're cited, where you're not, and what changed after the last model update.

  • ✓ Citation Rate: 81% (51/63 prompts)
  • ✓ Per-model breakdown (ChatGPT, Claude, Perplexity, Gemini)
  • ✓ Citation Delta tracking at 30/60/90 days
  • ✓ Framing analysis: positive, neutral, or negative citation
BM

Brand Memory

Your Brand Memory is the structured knowledge representation that AI models retrieve when processing prompts. We help you build, measure, and reinforce it continuously.

  • ✓ Brand Memory Score: 87/100
  • ✓ Entity assertions active: 42
  • ✓ Signal degradation alerts
  • ✓ Last model sync tracking
Learn more about Brand Memory →
EI

Entity Intelligence

Entity Intelligence maps, constructs, and weights brand entities for AI retrieval optimization. Entities are the atoms of AI knowledge — we make sure yours are defined, linked, and cited.

  • ✓ Entity definition and relationship mapping
  • ✓ Knowledge graph construction
  • ✓ Entity linkage density scoring
  • ✓ Cross-model entity persistence tracking
Learn more about Entity Intelligence →
CT

Competitive Citation Tracking

Know exactly who else is being cited for your prompts — and how your citation share compares. Track competitive movements across model updates.

  • ✓ Your brand vs. Competitor A vs. Competitor B
  • ✓ Citation share: You 81% / Comp A 12% / Comp B 7%
  • ✓ Competitive movement alerts
  • ✓ Gap analysis: where competitors are cited and you're not
Question Live search Answer with citations crawlability · passage clarity · freshness Perplexity — always ChatGPT — sometimes Gemini — sometimes Claude — sometimes Model-internal knowledge Answer without citations consistency across sources · category association ChatGPT — sometimes Gemini — sometimes Claude — sometimes Three platforms use both routes, which is why visibility is measured per platform.
Because the two routes reward different signals, a single visibility number hides which work is paying off. Which route a platform takes for a given question has never been published — so no ratio is claimed here, it is measured.
SA

Signal Scoring Engine

Our proprietary multi-dimension signal evaluation system that scores every potential citation signal by its actual retrieval impact. Not all signals are equal — the engine knows which ones matter. Published in full as open source.

Source Authority 25% Factual Consistency 20% Entity Linkage 15% Cross-Model Persistence 12% Temporal Freshness 12% Citation Frequency 10% Competitive Diff 6% total 100
The seven signals do not carry equal weight. The first three account for 60% of the score on their own, which is also the order the work follows.
Edge Functions

Built in-house.
40+ modules deep.

Every module — from Brand Memory to Citation Intelligence — was built by our engineering team. No white-label, no API wrapper.

Brand Memory Engine

Structured knowledge graphs that AI models retrieve and cite. Entity definitions, factual assertions, relationship maps, citation anchors — all built as machine-readable data.

42 entity assertions active

Knowledge Base

Curated, versioned, exportable repository of your brand's structured knowledge. JSON-LD export at any time. Portable to any system.

Full JSON-LD export

Citation Intelligence

Real-time citation rate tracking per prompt, per model. Framing analysis, Citation Delta at 30/60/90 days, per-model breakdowns.

756 tests per cycle

Prompt Discovery

Mine GSC, competitor citation data, LLM query patterns, and buyer journeys to build comprehensive prompt maps.

63 prompts discovered

Prompt Tracking

Continuous monitoring across all 4 models. Prompt-by-prompt citation presence with historical trendlines and anomaly detection.

4 models × 63 prompts

Content Gap Analyzer

Identify which prompts cite competitors but not you. Prioritized gap report with signal-level recommendations — not "write more."

Competitor gap scoring

AI Content Production

Entity-optimized content designed for AI retrieval. Structured for RAG systems, built for vector proximity, deployed across semantic surfaces.

Entity-structured output

Signal Scoring Engine

Multi-dimensional evaluation: authority (25%), consistency (20%), linkage (15%), persistence (12%), freshness (12%), frequency (10%), differentiation (6%). Published in whitepaper.

Open source algorithm

Competitive Citation Map

Real-time competitive intelligence. See who's cited for every prompt on every model. Track Citation Share over time.

You 81% / Comp A 12% / Comp B 7%

See the Platform in Action

Book a 15-minute walkthrough. We'll show you exactly how the platform tracks and improves your AI visibility — using your brand or a case study. Or read the published methodology first.

Book a Walkthrough →