en

Existing businessConsumer AI

Qaf-style cited Islamic Q&A copycat

Ask any Islamic question and get an answer researched across thousands of classical books, with openable citations — copied for Indonesian, Malay, Turkish and Urdu speakers.

2 weeks ago · Scored 25 Oct 2026 · ID HBV-261025-02

Working title: Qaf-style cited Islamic Q&A copycat for Bahasa Indonesia, Malay, Turkish and Urdu speakers (working name)

Summary

BriefingVerifiedHypothesis

Existing business benchmark. Qaf (qaf.ai) (https://qaf.ai/). Launch: Public launch post 5 Aug 2026 (X, Doomers/Launch Archive); an AI Daily article says it launched in May 2026 — disagreement logged, not resolved. Age: ~2.7 months since the 5 Aug launch post (up to ~5 months if May is right) (as of 25 Oct 2026).…

Read the rest

Verified (Doomers, read 6 Oct 2026): launch post 4,693,575 views. Verified (reference charts, 25 Oct 2026): entered DE, GB, US charts. Not found: users, downloads, revenue. X: 35 posts/7d (Up 29% Mon→Sat (full UTC days: Mon 19 7 → Sat 24 9); peak Sat 24 9; Sat 24 is 0% below peak; window 2026-10-18 11:00 UTC → 2026-10-25 10:00 UTC (X API 7-day limit; Sun 18 and Sun 25 are partial UTC days; full days Mon 19 → Sat 24)). Revenue hypothesis: ~$1K–$6K/month at month 6 (HYPOTHESIS).

Description

Ask any Islamic question and get an answer researched across thousands of classical books, with openable citations — copied for Indonesian, Malay, Turkish and Urdu speakers. What the benchmark does: Qaf searches the Shamela library of 8,000+ classical Islamic books, reads relevant passages and returns answers with openable citations (book, author, passage); it verifies Qur'an and hadith references and shows differing scholarly views.…

Read the rest

iOS, Android and web. Why replicable: Retrieval over a public-domain classical corpus with citations is a small RAG build; the language gap (Indonesia alone has ~240M Muslims) is real. But correctness risk is high and big players (Muslim Pro) already ship AI features. Monetization: Freemium subscription (unlimited questions, saved research), Ramadan premium packs; no ads on religious answers. Build: 2–3 agent-weeks (RAG over public-domain corpus, citations UI, ID/MS/TR/UR prompts); stack: Next.js on Vercel; Postgres + pgvector (Supabase); public-domain classical texts; LLM API; RevenueCat/Stripe. GTM: SEO/GEO pages per common question in each language; shareable cited answer cards; faceless short clips.

Target customer

Muslims in Indonesia, Malaysia, Turkey and Pakistan who want cited answers in their own language; diaspora in Europe.

Pain

Retrieval over a public-domain classical corpus with citations is a small RAG build; the language gap (Indonesia alone has ~240M Muslims) is real. But correctness risk is high and big players (Muslim Pro) already ship AI features.

Why this score

Why it scored lower: A wrong religious answer can destroy trust instantly; incumbents add the same feature.

MoatHypothesis

Weak–Medium: Corpus is public; citations UI is copyable. Scholar review and language quality could differentiate. Status: Hypothesis.

Demand signals

X posts (7 days)

35

Signal trend

Up 29% Mon→Sat (full UTC days: Mon 19 7 → Sat 24 9); peak Sat 24 9; Sat 24 is 0% below peak

Web demandVerified

Qaf Help Center (help.qaf.ai) explains the Shamela-based method; AI Daily (Greek) article; Launch Archive entry; LinkedIn launch post by the founder. Faith-AI category chatter 183 posts/7d, flat (Verified X API). Other X counts (same window, Verified X API): 183 posts: ("Bible Chat" OR "Christian AI" OR "catholic AI" OR "AI pastor" OR "AI priest" OR "Islamic AI" OR "AI imam") -is:retweet

Market and competitors

TAM

Not sized this run — Scope: Global excl. China. Method: No third-party report fetched. (Basis: n/a)

SAMHypothesis

Not sized — Scope: Target languages/markets above. Method: HBV assumption. (Basis: Hypothesis)

SOMHypothesis

~$12K–$72K ARR (hypothesis) — Scope: Month-6 run-rate. Method: From revenue hypothesis ×12. (Basis: Hypothesis)

Competitors

Muslim Pro (AI features), Ansari AI, Deen AI, IslamGPT-type apps, Bible Chat (Christian analog; HBV row HBV-261019-03 already covers a PT Bible Chat copycat).

Existing business

https://qaf.ai/

Business model

Monetization

Freemium subscription (unlimited questions, saved research), Ramadan premium packs; no ads on religious answers.

Revenue hypothesis (monthly)Hypothesis

~$1K–$6K/month at month 6 (HYPOTHESIS).

Traction and revenueVerified

Launch: Public launch post 5 Aug 2026 (X, Doomers/Launch Archive); an AI Daily article says it launched in May 2026 — disagreement logged, not resolved. Age: ~2.7 months since the 5 Aug launch post (up to ~5 months if May is right) (as of 25 Oct 2026). Verified (Doomers, read 6 Oct 2026): launch post 4,693,575 views.…

Read the rest

Verified (reference charts, 25 Oct 2026): entered DE, GB, US charts. Not found: users, downloads, revenue.

Jobs to be done

When I have a religious question and the answers online disagree, I want to get an answer in my language with the classical sources, so I can trust and share it with my study group.

HXC

Siti, 27, teacher in Bandung (illustrative) · Uses WhatsApp study groups; reads Bahasa Indonesia only. Buying signal: Asks religious questions online and checks sources; would pay a small subscription in Ramadan.

Build and go-to-market

Tech stack

Next.js on Vercel; Postgres + pgvector (Supabase); public-domain classical texts; LLM API; RevenueCat/Stripe

GTM channels

SEO/GEO pages per common question in each language; shareable cited answer cards; faceless short clips

Risks

Main risk

A wrong religious answer can destroy trust instantly; incumbents add the same feature.

All risksVerified

Value: Supported — 4.69M launch views (Verified, Doomers) | Usability: High — Religious accuracy | Feasibility: Attention — Crowded field | Business Viability: Attention — Flat X velocity

Next validation step

Do not build now. If Cesar wants a test: 100 cited Bahasa Indonesia answers reviewed by one scholar; measure organic visits in 14 days.

Sources

  1. qaf.ai
  2. help.qaf.ai
  3. launcharchive.ai
  4. ai-daily.gr
  5. doomers.ai
  6. X API counts (window above)

existingcopycatfaith-airagb2cindonesiaturkeypakistanglobal

Want the next one the day it is scored?

The score, the sources, the risks and the next step, by email.

Free. Sent with beehiiv. One click to unsubscribe.

  • New ideaConsumer AI· 3 weeks ago

    "É IA?" — forward any photo, video or audio…

    Forward a suspicious photo, video or voice note to a WhatsApp number and get, in Portuguese or Spanish, "likely AI / likely real / can't tell" with three simple reasons and a ready-to-paste reply for the family group. Aplicação global (any WhatsApp market); Versão LatAm first (Brazil, Mexico, Colombia, Argentina).

    Why it scored lower: Free incumbents and detector accuracy on WhatsApp-recompressed media.

    Open ProductBots.ai →

  • Existing businessConsumer AI· 3 weeks ago

    Muse-style proactive personal agent on…

    A personal agent that acts before you ask — reminders, bookings, bill follow-ups — delivered inside WhatsApp in Portuguese and Spanish, where Meta's Muse is not yet available.

    Why it scored lower: Meta launches Muse in Brazil (WhatsApp is Meta's) and the copy is erased overnight.

    Open ProductBots.ai →

  • Existing businessConsumer AI· 2 weeks ago

    Wabi-style copycat

    An AI agent that lives inside a group chat, answers everyone, plans trips and dinners, splits bills and spins up tiny shared apps — copied for WhatsApp groups in Brazil, Mexico and India.

    Why it scored lower: Platform risk: Meta restricts general-purpose AI bots on WhatsApp and ships its own group AI.

    Open ProductBots.ai →

Research, not investment advice. Numbers keep the label they were published with.