ن
noormap
verified directory of Muslim spaces

Roadmap

Global in intention, regional in execution. We are in Phase 0 (United States) and we have not launched outside it yet — by design.

Why we haven't launched globally yet

A directory of mosques that is wrong in dangerous ways is worse than no directory at all. The reasons we are not yet in your country are not technical timidity — they are concrete:

The temptation to scale prematurely is the failure mode of every Muslim institution. When we launch in your country, we want it to be the start of something durable — not a marketing date we have to walk back.

Phases

  1. phase 0
    United States
    Foundations
    Active
    Scope: 50 states
    Horizon: 2026 — in flight

    We start where the data is richest, where the regulatory environment is the most navigable, and where our team has direct reach. The US is not the prize — it is the proving ground. Every system, every rubric, every safety policy must work at one-country scale before we replicate it. The temptation to launch globally on day one is the failure mode of every Muslim institution. We resist it.

    In scope
    • OpenStreetMap US extract (1,223 mosques) + Ihsan Standard auditor (1,393 mosques with EIN + transparency claims)
    • HFSAA certified-entity directory (25 halal restaurants/butchers with full address + cert evidence)
    • Postgres + PostGIS schema with claim ledger; 420 cross-source duplicates auto-merged
    • UmmahPassport SSO with PKCE — confidential-client, pairwise sub, HMAC session
    • Privacy-first stack: MapLibre + OpenFreeMap + Esri satellite + self-hosted Photon (planned); coordinate snap to ~110m before any network call
    • Public rubric v0.1 — four pillars per entity type, partner-authored, semver-versioned
    QC gates · must pass to ship
    • Dedup precision ≥97%
      Cluster-and-merge pass on 150m radius matches by name similarity. We measured 420 / 2,641 records de-duplicated cleanly; sampled audit confirms <3% false-merges. The Valley Ranch case (OSM node + OSM polygon + Ihsan Standard) resolves to one canonical row.
    • Privacy contract upheld
      User coordinates never leave the device unsnapped. Server logs strip to 2 decimals (~1.1 km). Geocoder runs server-side so the user's IP never reaches Nominatim. Quarterly audit by an external HR org.
    • Defamation safety
      Hard rule: no negative glow on the map. Self-claim shortfalls are routed to the relevant cert body, not publicly red-marked. Cleared by counsel before any score goes live.
    • Methodology partner LOIs
      At least one signed LOI per scoring pillar before that pillar is enabled. Today: outreach to Bahu Trust (env), Lilly Ihsan Standard at IU (transparency), IFANCA + HFSAA + HMC (halal cert), ISPU (civic), Open My Mosque + Hind Makki (women's space), Janazah Project (funerals).
    Ground metrics · must hit before next phase
    • Audit coverage
      ≥250 mosques carrying at least one third-party-audited claim from a signed partner. ≥50 with full transparency-pillar coverage.
    • Cert-body coverage
      ≥500 halal restaurants with at least one active cert (HFSAA / IFANCA / HMC). ≥3 cert bodies actively syndicating their certified list.
    • Owner claim flow
      ≥100 mosques claimed by their leadership via UmmahPassport SSO. Edit audit log running cleanly. Takedown SLA <72 hours.
    • Contributor trust
      ≥1,000 UmmahPassport-verified contributors, weighted by historical claim accuracy. Paid editorial team in place for cold-start moderation.
    Gate to next phase

    Until every QC gate is passed and ground metrics show real adoption, we do not expand. A second country before the first is solid would mean propagating bugs, exporting our blind spots, and giving methodology partners we don't yet have license to speak for them.

  2. phase 1
    United Kingdom · Canada · Malaysia
    English-speaking diaspora + open halal state
    Next
    Scope: 3 countries
    Horizon: 2027 H1 — pending Phase 0 gates

    Three markets with strong open data and well-organized Muslim civic infrastructure. The UK gives us HMC / HFA / Halal Food Authority + MCB's Women in Mosques work. Canada gives us CCMW + open government data. Malaysia gives us JAKIM — the gold standard of state-led halal certification. We replicate the US playbook with country-specific certifier and registry adapters; no architectural changes required.

    In scope
    • Country adapters: HMC + HFA (UK), HMA (Canada), JAKIM MyeHalal (Malaysia)
    • Mosque sources: MCB UK directory, NCCM Canada, OSM expansions in all three
    • Women's prayer-space rubric live (Open My Mosque "My British Mosque" methodology, co-published)
    • Per-country sect-tag taxonomy and inclusion-policy localization (e.g. Ahmadiyya in UK is a load-bearing case)
    • Local advisory councils convened before any per-country rubric goes live
    QC gates · must pass to ship
    • Per-country defamation review
      Counsel review in each jurisdiction before scoring publishes. UK's Defamation Act 2013 has different defenses than US Section 230 reasoning.
    • Localized inclusion policy
      Sect taxonomy reviewed by local scholars per country. Public editorial policy notes per-country deviations and the reasoning.
    • Cert-body partnership signed
      At least one local cert body has signed a data-sharing or attribution agreement before launch. We do not scrape adversarially.
    Ground metrics · must hit before next phase
    • Ecosystem density
      ≥80% of self-identifying mosques in each country present in the directory at launch.
    • Women's-space coverage
      ≥200 mosques per country with a Side Entrance / Open My Mosque-style rating against the published rubric.
    • Halal cert verification
      ≥1,000 restaurants per country with at least one active cert, surfaced with the rigor-of-cert label.
    Gate to next phase

    Each country has its own QC + ground-metric gates. We may launch any one of the three before the others if its gates pass first. We do not bundle.

  3. phase 2
    Indonesia · Singapore · Turkey · GCC
    Muslim-majority + state halal infrastructure
    Future
    Scope: Government-led ingest
    Horizon: 2028 — pending Phase 1 gates

    Markets where the state has done much of the work for us — Diyanet's 85,000-mosque registry in Turkey, BPJPH's mandatory SiHalal system in Indonesia, MUIS in Singapore, the Awqaf ministries in Saudi Arabia and the UAE. Our job here is to be a respectful surface for data the state already publishes, not to duplicate or compete. Partnership-led, not scrape-led.

    In scope
    • Diyanet mosque registry (Turkey) — government-to-org agreement required
    • BPJPH / MUI / LPPOM (Indonesia) — mandatory state halal system
    • MUIS (Singapore), Awqaf MoIA (Saudi), UAE Awqaf
    • Multilingual UI (Bahasa Indonesia, Türkçe, Arabic — RTL-ready)
    • Per-country governance councils — no rubric ports without local sign-off
    QC gates · must pass to ship
    • Government data-use agreement
      Formal MoU with the relevant ministry before launch. No scraped use of state-published mosque lists without permission.
    • RTL + script coverage
      Arabic, Urdu, Bahasa Melayu rendering verified end-to-end including the entity card, rubric explanations, and the claim ledger.
    • Currency + locale parity
      Distance units, date systems (Gregorian + Hijri toggle), and prayer-time defaults localized.
    Ground metrics · must hit before next phase
    • State registry adoption
      ≥3 governments syndicating their mosque or cert registries directly to noormap, with attribution and update SLAs.
    • Native-language contributors
      ≥5,000 UmmahPassport-verified contributors operating in their native language, with moderation in-language.
    Gate to next phase

    Government partnerships are slow work. We do not pre-announce a launch date. We announce when an MoU is signed.

  4. phase 3
    Sub-Saharan Africa · South Asia · Latin America
    Long-tail discovery
    Future
    Scope: Community-led, partnership-anchored
    Horizon: 2029+ — pending Phase 2 gates

    The regions where OSM is patchy, state registries are absent or unsafe, and the right partners are local NGOs and chapter networks. The unit of ingest here is not the country but the city or the federation. Lagos, Karachi, Hyderabad, Dakar, Buenos Aires — each with its own anchor partner. Slowest phase, highest dignity stakes, lowest tolerance for getting it wrong.

    In scope
    • SMS "nearest mosque" lookup via Africa's Talking — for users without smartphones or steady data
    • PWA + offline country packs — download a country, work offline for a week
    • Per-city anchor partnerships rather than per-country government deals
    • Cemetery dignity-gating layer enabled where local partners have signed the ToS
    • Native-language interfaces (Hausa, Wolof, Urdu, Bengali, Swahili, Spanish)
    QC gates · must pass to ship
    • Local partner present
      Every city goes live with at least one named anchor partner. No drive-by data ingest without a community contact.
    • Low-bandwidth verified
      Map + list view tested on $40 Android over 2G/3G. <500 KB initial payload. Offline pack download <50 MB per country.
    • Family-grave dignity ToS
      Per-cemetery agreement covers next-of-kin verification, opt-in display, GDPR-style erasure regardless of jurisdiction.
    Ground metrics · must hit before next phase
    • SMS adoption pilot
      ≥10,000 SMS lookups/month in two African markets within 6 months of launch.
    • Native moderators
      ≥1 paid native-speaker moderator per launch city for the first 12 months.
    Gate to next phase

    These markets reward patience and punish ambition. We do not export the US playbook here. We co-design the playbook with local partners or we don't launch.

  5. phase 4
    High-surveillance regions
    Suppressed-by-design
    Conditional
    Scope: Region-level only
    Horizon: Indefinite — conditional on local safety

    Xinjiang, parts of India under CAA/NRC tension, Burma, and any other region designated high-surveillance by partner human-rights organizations. In these regions noormap suppresses precise location, contributor identity, attendance signals, and per-entity detail. Map shows region-level aggregate only. Contributors operate pseudonymously even though identity is still UmmahPassport-verified. This phase exists because the alternative — pretending these regions aren't there, or showing them with the same fidelity as everywhere else — both fail the safety test.

    In scope
    • Region list curated by partner HR orgs (Uyghur Human Rights Project, Justice for All, regional equivalents)
    • Aggregate-only rendering: count and approximate area, no individual entity dots
    • Pseudonymous contributor flow with UmmahPassport-verified identity backstop
    • External annual audit of the suppression policy
    • Public appeals process for the region list — additions, removals, severity
    QC gates · must pass to ship
    • HR-org sign-off
      Region list updated only after explicit sign-off from at least two independent partner organizations. Public diff of every change.
    • Suppression integrity
      External penetration test confirms no contributor identity, no precise location, and no attendance-pattern leakage from the aggregate view.
    Ground metrics · must hit before next phase
    • Trusted by affected community
      Direct feedback from community organizations in each region confirming the suppression model serves them rather than abandoning them.
    Gate to next phase

    No exit from this phase. A region either remains here or graduates to a lower-risk phase based on partner advice — never on our own assessment.

How you can speed this up

  • If you run a mosque or cert body: sign in via UmmahPassport and claim your entry. Owner-verified records jump tiers instantly.
  • If you are a methodology partner (a women's-space body, an environmental auditor, a halal cert body, a janāzah service network): email partners@noormap.com — we co-author the rubric in your category, you own it, we host the data.
  • If you have local-language moderation capacity in a Phase-2 or Phase-3 region, we want to talk before that region launches.
  • If you have community context about a region we should add to the high-surveillance list — or remove from it — write us. The list is co-curated.

See also: the map · hello@noormap.com

Focus area

Prayer-times coverage

End users come to NoorMap for two things first: is this mosque real / verified and when does it pray. Iqamah times are inherently mosque-set — they shift with the seasons, with Ramadan, with special events, sometimes weekly — so the work is never “done.” Here’s how we’re building toward comprehensive, fresh, mosque-controlled coverage.

  1. Phase 1 · Live
    2026

    Surface what mosques already publish

    • Fetch and render iqamah live from Athan+ by MasjidAl for every mosque whose masjid_idwe’ve discovered
    • Refetch every hour at minimum; show “last verified · X min ago” on every page so users can spot stale data at a glance
    • Astronomical fallback (ISNA convention) when we have no iqamah — clearly labelled prayer windows, not jamāʿah times
    • Common-Crawl harvest of every publicly-embedded Athan+ widget on the web → discovers mosques globally without asking each admin
  2. Phase 2 · Next
    2026 H2

    Mosque admins claim & publish on our stack

    • Sign in via UmmahPassport with the org_rep scope; verify mosque admin via domain email, phone OTP, or document review
    • Once verified, three options:
      • link your existing Athan+ masjid_id — we keep using their live widget
      • link an existing prayer-times page on your website — our scraper keeps it in sync
      • enter your iqamah directly on Ihsan Standard (free, no third-party dependency)
    • Seasonal schedules: default + Ramadan + weekly Jumuʿah + ad-hoc overrides (Eid, janāzah)
    • Times entered by a verified admin override every other source on the listing
  3. Phase 3 · Following
    2027

    NoorMap-native widget for mosque websites

    • Embeddable widget (iframe + JS embed) mosques can paste into their website to display their iqamah — same shape as Athan+, no tracking, no ads
    • Themeable; respects mosque’s typography + colors
    • Single source of truth: the iqamah a mosque sets on Ihsan Standard is what appears in their own website widget AND on NoorMap AND in any future NoorMap app
    • Frees mosques from third-party dependency without forcing them to leave Athan+ — both surface the same underlying schedule via the admin’s choice
  4. Phase 4 · Coverage at the edges
    ongoing

    Community-tipped + heuristic-scraped coverage

    • Any signed-in UP user can submit a tip: “mosque X publishes prayer times at URL Y.” Two independent tips → auto-add to the scrape queue
    • Heuristic prayer-times parser that handles common WordPress + Wix prayer-time blocks (the long tail of mosques that publish on their site but use no third-party widget)
    • “Report incorrect” on every prayer-times card feeds a moderation queue; verified admins resolve

Freshness contract: we never show stale prayer times silently. If our last successful fetch was more than 24 hours ago, we surface that on the card itself with the timestamp. If it’s more than 7 days ago, we hide the times entirely and revert to the astronomical fallback + a “our last sync failed — help us re-link” CTA. Trust requires this.

Mosque administrators: email listings@ihsanstandard.org to claim early access to Phase 2.