Your experts. Your knowledge.
Verified by Socratic dialogue.
FedWell turns expert conversations into searchable, citable, evidence-graded knowledge bases. No more wikis that rot. No more PDFs that hide. Just living knowledge that gets better every time an expert talks.
Socratic Dialogue Engine
An AI interviewer that doesn't nod along. It challenges claims, demands citations, and only signs off when the evidence holds up. Think of it as peer review, but faster and less passive-aggressive.
Evidence-Graded Claims
Every consensus point gets a clear evidence level — from meta-analysis down to expert opinion. No more guessing whether a claim is based on a landmark trial or someone's vibes.
Semantic Search
Your knowledge base understands meaning, not just keywords. Ask it a question in natural language and it'll find the right consensus point, even if the wording is completely different.
How It Works
Same process that powers fedwell.ai, white-labelled for you.
Invite Experts
Add your team — clinicians, researchers, engineers, whoever holds the knowledge.
They Talk
Experts have Socratic dialogues. The AI challenges, probes, and refines every claim.
Knowledge Forms
Verified claims become consensus points with evidence levels, sources, and attribution.
Everyone Benefits
Embed, search, or share. Your knowledge base is alive and always improving.
Who It's For
Healthcare Organisations
Turn your clinical expertise into a verified knowledge base that staff, patients, and partners can actually search and trust. No more 200-page clinical guidelines that nobody reads.
"What does our evidence actually say about X?"Research Communities
Capture the consensus in your field as it forms, not after five years of committee meetings. Track contested claims, evidence levels, and who contributed what.
"Where do we actually agree, and where don't we?"Professional Networks
Give your members a knowledge base that reflects collective expertise, not just the loudest voice. Every claim earns its place through evidence, not seniority.
"What do our best practitioners actually recommend?"Built-In Trust Mechanisms
Every domain has its own relationship with evidence. FedWell adapts.
Flexible Evidence Levels
Not every valuable claim comes with a citation. FedWell supports nine evidence levels — from meta-analyses to experiential knowledge and institutional data. A veteran litigator's "30 years of practice" gets its own badge, honestly labelled, rather than being forced into an academic framework or rejected entirely. The Socratic dialogue adapts its questions accordingly: published research gets challenged on methodology; experiential claims get probed for consistency, counter-examples, and peer agreement.
Peer Corroboration
When an expert stores a consensus point, the system can request corroboration from another expert in the panel — chosen at random to avoid confirmation bias. That expert can confirm the claim, dispute it (triggering a visible "contested" flag), or return it with a suggestion for refinement. This lightweight peer review catches blind spots without creating a bureaucratic bottleneck.
12-Month Review Cycle
Knowledge isn't static. Every consensus point carries a review-due date, 12 months from creation. When the date arrives, the contributing expert is invited to re-examine their claim in light of new evidence. Points can be reaffirmed, updated, or retired. No more zombie guidelines that everyone follows but nobody maintains.
Visible Debate Threads
When experts disagree, the disagreement becomes a feature, not a bug. Contested points display both perspectives with full attribution. For deeper disputes, experts can host a recorded video discussion that lives right alongside the contested claim — a civil, structured conversation that readers can watch and judge for themselves.
Expert Panel Curation
FedWell helps you build a balanced panel by tracking diversity across multiple axes: institution type, geography, career stage, and theoretical orientation. The system flags when your panel is skewed and suggests adversarial appointments — experts known to hold minority positions who keep the consensus honest. Experts serve staggered terms to prevent groupthink.
Explicit Evidence Confirmation
The Socratic dialogue doesn't just label evidence levels behind the scenes. Before any consensus point is proposed, the AI and the expert explicitly agree on how to classify the evidence. "I'd call this professional consensus rather than an RCT — do you agree?" This ensures the expert owns the classification, and readers can trust the label.
Interested?
We're rolling this out thoughtfully. If you think your organisation's knowledge deserves better than a shared drive and good intentions, let's talk.
Get in touch