Brand Knowledge is Redsun's evidence pipeline: you add artifacts, sales-call and pitch transcripts, case studies, tone-of-voice documents, and Redsun combines them with your live CRM data to draft proof points, buying triggers and a voice profile. You confirm every claim before it can appear in your website copy, so AI-written pages stay grounded in verifiable evidence.
More context in, a sharper voice out
A worked example on Ledgerline Search, an illustrative fintech agency: the same two lines redrafted three times, each pass grounded in more of the agency's own context.
Worked example. Nothing on this page analyses your site; the real extraction runs on your own site at onboarding.
The scan reads what the site already says
Paste a URL at onboarding and Redsun reads the pages: sectors, services, stats, testimonials, CTA language, colours and fonts. The first voice profile is accurate and thin, because a site can only give back what it already says.
- ledgerline.example14 pages read at onboarding
- Site copy mined2 stats, 3 testimonials, CTA language
- Paymentssite copy
- Lendingsite copy
- Fintech operationssite copy
- Placing fintech teams since 2014site copy
Empty. A veto list needs the agency's own words.
Specialist fintech recruitment across the UK and Ireland.
We're hiring across payments and lending. Our latest roles are on the site.
Drafts follow the same rules the copy engine enforces for every tenant: numbers only from records, claims only from confirmed evidence, and the banned-phrase list applied throughout.
Two sources of evidence. One standard of proof.
Redsun draws a hard line between facts it can read straight from your data and judgment calls it drafts from your words. They are handled differently on purpose.
- Placement counts and fill times, from real records
- Priority roles, from your live jobs
- Sectors and locations you actually work
- Best-fit clients and buying triggers, from your material
- Proof points, each citing the source it came from
- A voice sample distilled from how you actually pitch
Nothing here reaches your site until you confirm it. AI drafts; you decide.
You are the editor. Try it.
Every drafted claim lands here first. These are Ledgerline's, from the worked example above. Confirm what's true, edit what's close, reject what isn't. Only confirmed evidence can feed your copy.
Average time-to-fill of 18 days across fintech engineering roles
Derived from 41 placements in your CRM (last 12 months)
Trusted by Series-A fintechs scaling their first engineering teams
Cited from a sales-call transcript (2 Feb)
Specialists in regulated payments and lending hiring
Cited from your capability statement + case studies
These describe how the pipeline works, not a performance benchmark: CRM facts are drafted deterministically, AI suggestions are quarantined, and no claim is published until you confirm it.
The antidote to generic AI copy
Your voice, from real conversations
Transcripts are mined for how your consultants actually pitch, the phrases, the mechanisms, the objection handling, and distilled into a voice sample that steers every generated page.
Evidence quality is the AEO gate
AI engines cite content they can verify. Because every claim on your site traces to a confirmed proof point or a real CRM record, your pages carry the evidence signals answer engines reward.
Grounded, retrievable knowledge
Confirmed knowledge is embedded for retrieval, so Helion and the copy engines pull from your actual material rather than guessing from generic recruitment tropes.
It also learns what NOT to say
The evidence layer includes claims-to-avoid, conservative boundaries drafted from your material, so generated copy never overpromises on your behalf.
Brand Knowledge FAQ
Common questions about Redsun's evidence and brand-voice pipeline.
- What is Brand Knowledge in Redsun? +
- Brand Knowledge is Redsun's evidence pipeline, in the Brand hub. You upload artifacts — sales-call and pitch transcripts, case studies, tone-of-voice documents, proposals — and Redsun combines them with your CRM data to draft proof points, buying triggers, best-fit client profiles and a voice sample. Every claim sits unconfirmed until you approve it, and only confirmed evidence can feed your website copy.
- How does Redsun learn my agency's tone of voice? +
- The highest-trust source is your real conversations: sales-call and pitch transcripts are cleaned (speaker labels kept, so the system knows which voice is the agency's) and distilled into a voice sample, alongside tone-of-voice documents and existing content you upload. Every AI-written page then follows that voice rather than generic recruitment copy.
- Does AI invent facts about my agency? +
- No. Hard facts — placement counts, fill times, priority roles — are drafted deterministically from real CRM records, never by AI. AI-drafted suggestions (proof points, triggers, positioning) are grounded only in the material you provided, cite their source, and sit in a quarantine until you confirm them. Only confirmed evidence can appear in copy.
- Why does evidence matter for AEO? +
- AI engines like ChatGPT and Google AI Overviews favour content they can verify. Copy that traces to confirmed proof points and real CRM data carries the evidence signals answer engines reward — which is why evidence quality is the gate in Redsun's AEO scorecard.
- What happens if I never add any artifacts? +
- Redsun still generates your site from your onboarding context, sector data and CRM connection. Brand Knowledge is additive: the more evidence you confirm, the more specific, provable and on-voice your copy becomes — and you can add artifacts at any time.