Shaman AI
Shaman’s AI is built for pharma content authoring. It’s woven into every step of how your brand teams and content studios already work — search, write, translate, tag — and grounded in the approved claims, references and mandatory statements that live in your Veeva Vault PromoMats. Aurora, our agentic AI, composes review-ready assets from that single source of truth, with provenance on every element. A glass box, not a black box.
AI in every step of authoring
AI woven into the authoring flow, so a compliant asset comes together in minutes instead of days.
Localise content into any market language in minutes — including visual components, not just text — so global-to-local stops waiting on an agency round-trip.
Semantic search and find across your whole organisation’s approved content. Ask for a claim or a concept and surface every approved place it already lives.
GenAI writing built into the editor: draft from a brief, rewrite a passage, or summarise long source material — always from approved wording, never beyond it.
AI that does the busywork: find the right images, generate alt text, and auto-tag content so every asset is classified and accessible without manual effort.
Aurora · agentic AI
Aurora goes beyond assistance. Give it an approved source from Vault — a detail aid, a set of claims, a study summary — and it composes new assets for email, CLM, web or print, assembling from approved modules and preserving the link to every claim and reference back to PromoMats. Its specialised agents understand medical content, editorial craft, your brand design system and context, coordinated by an orchestrator.
Why it is safe
Shaman’s AI draws only on approved, MLR-cleared knowledge from Veeva Vault PromoMats — your single source of truth — so every element it produces carries provenance back to an approved source. That’s the difference with a general model, which writes fluent language with no notion of what your company is approved to say. It runs on the same foundation as the rest of the platform: Atlas grounds it in your approved Vault content, Authoring is where you put it to work, and MLR review stays fast because every element already traces to something Vault has approved. AI as a compounding advantage — not a compliance risk.
FAQ
A general LLM generates fluent language from patterns in its training data — it has no concept of what your company is approved to claim. Shaman's AI is grounded in your approved content: claims, references and mandatory statements that have already cleared MLR. Every element it produces traces back to an approved source.
No. It assembles from approved modules and cited references. Where it drafts new phrasing, the underlying claim and its reference remain linked, so reviewers see exactly what's supported and what's new.
Aurora works on top of your approved content in PromoMats. Modules, claims and references stay governed in Vault; Aurora composes review-ready assets from them and syncs back for MLR.
No. Shaman never uses your content or data to train AI models. Your approved content grounds generation for your team only — it is used to compose your own assets, never to train a shared or third-party model.
Email, CLM, web and print — assembled from the same approved modules.
Trusted by the world’s leading life sciences organisations
Discover how to produce better content, faster — and take back control.
