Qorpera
Run AI Analysis
Download Qorpera

Platform

OverviewHow it worksGetting startedThe wiki

Operations

SituationsReasoning engineTrust gradientProjectsSystem jobs

Integrations

Connecting toolsGoogle WorkspaceMicrosoft 365SlackHubSpotStripe

Governance

Policy engineAudit trailSecurity & compliance

Platform

The wiki

The wiki is Qorpera’s single source of truth. Unlike retrieval systems that re-derive understanding from raw data on every query, or traditional databases that store flat records, the wiki is a persistent, LLM-maintained synthesis of organisational intelligence.

The wiki IS the system. A detected situation is a page. Its investigation, action plan, step results, and timeline are sections on that page. The execution engine reads and writes steps directly. The frontend renders directly from page content.

Page types

Operator-scoped pages (your company’s knowledge):

  • situation_instance — a detected situation’s lifecycle: trigger, investigation, action plan, timeline, deliverables, learnings
  • situation_type — detection criteria, playbook, learned patterns
  • person_profile — team member with activity stream
  • domain_hub — department or division with activity stream
  • company_overview — top-level organisational page
  • external_contact — person outside the organisation
  • external_relationship — company, vendor, or client relationship
  • process — business process documentation
  • project — active project with deliverables
  • initiative — proposed improvement or strategic idea
  • tool_system — internal tool or system documentation
  • system_job — autonomous background work agent

Plus system-scoped pages for the practitioner-reference layer: tacit_knowledge, benchmarks, red_flags, decision_heuristics, regional_practice, and pattern_recognition.

Trust levels

Every page is versioned with a trust level that gates how heavily it influences reasoning:

  • Provisional — newly created, unverified
  • Established — verified by independent model
  • Authoritative — consistently cited in approved reasoning
  • Challenged — contradicted by newer evidence
  • Quarantined — failed verification, excluded from reasoning until re-synthesised

This biological defence architecture prevents errors from compounding — every page is independently verified before it can influence decisions.

Cross-references

Pages link to each other through [[wiki-slug]] syntax. The reasoning engine follows these links during investigation, navigating from overview pages to specific expertise as needed. Hub pages serve as domain entry points; leaf pages contain deep expertise and cross-reference freely across domains.

Activity streams

Person and domain pages carry activity streams — temporal records of connector-sourced signals (emails sent, meetings held, Slack threads participated in). Streams are the raw material for behavioural pattern analysis, and they compose with wiki context during situation reasoning.

Previous← Getting startedNextSituations →

AI that actually works inside your business.

Product

  • Qorpera Cloud
  • Qorpera Desktop
  • Data & hosting
  • AI Check

Learn

  • Tutorials
  • FAQ
  • Research
  • News
  • Use cases
  • AI workforce integration
  • Company context for AI
  • Supervised AI execution
  • Documentation
  • About

Company

  • Contact
  • Private beta

Legal

  • Privacy Policy
  • Terms of Service

© 2026 Qorpera ApS · Copenhagen, Denmark · contact@qorpera.com

Data stored in EU (Frankfurt) · Encrypted with AES-256-GCM