CEDA
Governed AI-readiness diagnostics for professional-services firms.
CEDA (Controlled Evolution Diagnostic Application) is an AI-powered diagnostic platform for management consultants assessing AI readiness in professional-services firms. It turns unstructured consulting conversations into a governed, repeatable discovery process — surfacing operational fragility, knowledge-concentration risks, and AI opportunities.
It produces executive-grade strategic reports in a fraction of the time a manual audit requires, making recommendations only where operational architecture supports them and regulatory constraints permit.
Structure before augmentation. Governance before automation.
System Capabilities
What it does
Interview Conductor
Six-stage structured discovery powered by Claude. Real-time entity extraction, per-stage scoring, practitioner override, multi-session pause/resume.
Voice Interview Mode
A conversational audio layer: the client speaks via browser mic, CEDA listens and responds with synthesised speech. Silence-based turn-taking and barge-in.
Client Self-Service Portal
Public interview via secure time-limited links — no account. Clients complete the full six stages autonomously; practitioners review async.
SOP Topic Framework
Organise interviews around SOP topics with multiple respondents; consolidated views detect contradictions, gaps, and synthesis patterns.
AI Readiness Assessment
Five-dimension maturity scoring with floor constraints, radar visualisation, and a practitioner review checkpoint with lock-in.
Compliance & Ethics Filter
Hybrid screening: Claude classification plus a 26-rule deterministic EU AI Act engine, RAG-enriched, with composite risk scoring and full audit trail.
Feasibility Scorer
Five-dimension opportunity evaluation with readiness-ceiling adjustment, so scores reflect actual firm capability and map to actionable recommendations.
Document Ingestion & Evidence Layer
Upload SOPs, process docs and org charts (PDF, DOCX, XLSX, PPTX, images) with automated quality scoring and drift detection against interview-extracted reality.
Report Generation & Approval
Three executive deliverables — Opportunity Assessment, Risk-Impact Matrix, Implementation Roadmap — with per-section review, locking, and selective regeneration.
Roadmap Registry & Progress Tracking
Roadmap items auto-seeded from approved opportunities and readiness gaps, with status lifecycle, evidence logging, and readiness re-scoring over time.
Knowledge Base
Version-controlled regulatory and ethical framework library with jurisdiction-specific overlays, feeding the compliance filter and feasibility scorer via RAG.
Multi-Tenant Engagement Management
Engagements per client with firm context, multi-session support, lifecycle tracking, and role-based access across practitioner tenants.
The Process
6 structured stages
Diagnostic
Interview
Analyse
Roadmap
Implement
Monitor
Heritage & Authority