# Project context for AI: Iqrar — project context

Cryptographically verifiable AI compliance evidence — multi-authority quorum-signed bundles for supervisors and regulated firms. Read before working on or evaluating Iqrar. https://iqrar.io

## What this project is

Iqrar is the cryptographic evidence layer for AI compliance: multi-authority quorum-signed rule bundles, regulator-issued runtime directives, a committed audit trail anchored on Sigstore Rekor, and source-boundary channel separation of agent telemetry — verifiable end-to-end without trusting Iqrar infrastructure. Provisional patent PT 2026/04927 filed. Spun out of Stratafy.ai.

## Strategic thesis

AI agents act in milliseconds; their supervisors see quarterly filings. That gap is where every serious AI risk lives. Iqrar closes it — regulators push binding directives and verify compliance cryptographically; firms prove compliance without opening their infrastructure. No firm operates in a single regulator's silo, so the architecture is multi-authority from day one — and single-regulator products cannot retrofit that cleanly.

## Out of scope (do NOT build)

- General-purpose policy enforcement (OPA/Rego) — Iqrar is specifically regulator-issued runtime directives, not an arbitrary policy engine
- OpenTelemetry replacement — Iqrar telemetry is compliance-oriented, channel-separated, and regulator-queryable; not a metrics/tracing layer
- Single-operator trust models — the point is that no single party, including Iqrar itself, needs to be trusted

## Reasoning anchors

- Verification without trust — every claim is cryptographically verifiable end-to-end without trusting Iqrar infrastructure
- Channel separation at source — telemetry encrypted at the SDK boundary; the routing worker cannot read plaintext
- Foundation governance over any single operator — registry updates require threshold multisig
- Regulator-first design — directives, challenges, and audit proofs are designed around regulator workflows, not operator convenience

## Tone

Direct and technical. Audience is regulators, compliance engineers, and AI agent builders — not a general developer audience.

## Url

https://iqrar.io

## Positioning

- SupTech provider to regulators — not itself a regulated financial service
- Clean-anchor strategy: prove the architecture in single-coherent-regulator regimes (DIFC/DFSA-type), validate in messy multi-authority regimes (SA: POPIA / FSCA / NCR / FICA)
- Supervisor parity as the product: what an agent does in milliseconds, its supervisor can verify in seconds

## Partner Profile

- **Regulatory Suptech:** Financial supervisors with AI agent governance on the roadmap — not a vendor relationship, a co-design relationship. Specifically: innovation teams at single-authority regulators (DFSA/DIFC-type) who are piloting machine-readable supervisory frameworks or continuous compliance monitoring. Multi-authority regimes (SA, EU) follow once the clean-anchor case is established.
- **Regulated Firms:** AI-first fintechs, banks, or model providers deploying agents across 2+ authorities who currently produce manual compliance evidence. The question they're asking: 'how do I prove to four regulators simultaneously that my agents did what they were supposed to?' — not 'how do I run policies.'
- **Protocol And Ecosystem:** AI agent framework builders whose users face compliance questions (LangChain, Mastra, etc.) — SDK integration opportunity. W3C VC / Sigstore / IETF participants for protocol peer review and standards positioning.

## How To Engage

- Regulatory innovation teams — reach out with a specific pilot framing; we'll co-design the supervisory interface around your workflow
- Regulated firms — start with the three-line SDK integration on a non-production agent; we'll scope a proof-of-compliance run together
- SDK / framework integrators and protocol peers — contact routes at https://iqrar.io
- Start at https://iqrar.io


## Other lenses

- **Stratafy — project context** (project-context) — Paste into an AI tool working on or evaluating Stratafy. What it is, who it's for, current stage, and how to engage.
  https://aetherid.ai/leonard-cremer/stratafy
- **AetherID — project context** (project-context) — Paste into an AI tool working on or evaluating AetherID. What the protocol is, current state, scope discipline, and how to engage.
  https://aetherid.ai/leonard-cremer/aetherid

---
Source: https://aetherid.ai/leonard-cremer/iqrar
Owner profile (connected accounts, verified claims): https://aetherid.ai/leonard-cremer
Updated: 12 June 2026