Skip to main content

Methodology

Evidence model and technical boundaries

Plain-language detail for procurement, IT, and evaluation leads — aligned with the public truth table.

This page describes how CIIS is designed to behave. Capability-by-capability live status is maintained on current status (last reviewed June 2026).

Evidence model

  • Outcome statements link to measures, source evidence, review state, and explicit limitations.
  • Shareability labels (ready to share, internal only, needs review, blocked) reflect evidence strength and governance — not marketing tiers.
  • Reporting lenses map interpretation for different audiences; they do not certify impact or compliance.

Human review gates

  • CIIS supports judgement; it does not replace it.
  • No automated funding, clinical, or policy decisions.
  • Reviewer context and next actions are part of the record, not optional footnotes.

Data ingestion (today)

  • CSV-first bulk import for spreadsheet-style program data.
  • Document and manual-entry workflows are pilot-scoped.
  • Native multi-sheet Excel import is not supported today — see current status.
  • APIs may be enabled by agreement; not open public API self-serve.

AI assistance

  • AI may assist drafting, classification, or quality signals where enabled in scope.
  • Outputs are advisory; accountable humans approve what leaves the organisation.
  • The public homepage claim demo uses fixed illustrative rules — not production models or your data.

Privacy and deployment

  • Your organisation owns the impact information it enters; exact subprocessors and retention are contract-specific.
  • CIIS targets aggregate outcomes and performance information, not individual case management — minimisation remains a shared responsibility.
  • Compliance posture (e.g. APP, GDPR-oriented workflows) depends on configuration and your processes — not software alone.

What we do not claim

  • Fixed counts of quality checks or shipped report types beyond what current status lists.
  • ROI or time-saved at scale without measured pilot evidence.
  • Open self-serve product access without a guided pilot conversation.

Relationship to Impact Intelligence

Impact Intelligence Lab builds CIIS for program-level evidence governance and reporting. Impact Intelligence (impactintel.com) is a separate consultancy focused on large-scale unstructured data analytics for global institutions. The Lab site does not describe news-corpus or enterprise strategic-intelligence engagements; those belong on impactintel.com.

Questions for IT or procurement?

Request a governance briefing or pilot scope conversation with technical detail.

Test your weakest claim