Ask610
It started with Read.
Sunayna Agha, Founder · Stratenow LLC (New York) · [email protected] · ask610.com · ai.ask610.com · Ask610 on X
October 2026
The questions Muslims carry alone
“Is there any repentance for me?”
The question people carry for years, and no one to ask.
Religious guidance was once carefully handed down through trusted sources. Today, digital platforms accelerate the spread of information, prioritizing speed over depth. Fatwa sites answer from one school. AI invents citations. This shift makes it harder to verify sources and recognize the diversity of scholarly opinion.
Fourteen centuries of scholarship sit digitized in Arabic. Today’s apps are built either for scholars or for beginners, rarely both, and few tell you when the evidence behind an answer is thin.
Competitive landscape: everyone shares the same goal. Here’s what only Ask610 brings together.
In one product: Simple and Scholarly modes · four-school comparison in Scholarly mode · scored answers, named scholar governance · answers on X
| Player | What they do well | What they don’t have |
|---|---|---|
| Usul.ai (nonprofit) | Scholarly mode: scholar-grade search, 15,000+ classical texts; $5 / $25 / $100 | Simple mode. Four-school view. Scored answers. Cited answers on X |
| Qaf (for profit) | Simple mode: fast consumer chat over classical texts, 14 languages; $19.99/mo | Scholarly mode. Four-school view. Scholar governance. Scored answers |
| Ansari (open source, free) | Free, 140K+ conversations, publishes its own evaluation | Four-school view. Named review. Per-answer scoring. Cited answers on X |
| Islamify (nonprofit, Munich) | Four schools by default; scholars’ council; EUR 5-20/mo | Named scholars. Scholarly mode. Classical-scholar citations. Cited answers on X |
| AI Fiqh (Malaysia) | Fiqh and Islamic-finance corpus; User and Scholar versions; RM 29-200/mo | Aqīdah, Sīra, contemporary scope. Review board. Scored answers. Cited answers on X |
| Dalail (Saudi waqf, beta) | Four schools with each school’s evidence; Arabic-first | English. Review board. Scored answers. Cited answers on X |
Bottom line: This market isn’t saturated. It’s early innings for trusted Islamic AI, and Ask610 is the only one putting all four in one product.
Solution thesis: what a trusted Islamic AI is and is not
- Learn the basics of worship, step by step
- Find a half-remembered verse, Ḥadīth, or duʿāʾ, with Arabic and context
- Study the Sīra and the human side of faith: anger, grief, repentance
- Clear away fabricated stories and cultural myths
- Walk into your imam’s office with a sharper question
Learn here. Verify at the source. Take personal rulings to a living scholar.
- Search across a classical corpus of 8,425 books and a contemporary corpus from Prophetic Guidance
- Side-by-side school positions with citations, live
- Research Mode: chain tracing and drafting help, on the roadmap
AI: the next vessel for wisdom. The charge to honor and protect knowledge remains unchanged. The scholar remains the final authority.
Scholars govern the system: they gate the corpus; every answer is scored on a six-axis protocol with hard fails; and scholars work the review queue, failures first. When sources run thin, the answer says so.
On one screen
One question. Every position cited.
Same engine, pointed at social media.
Two retrieval lanes (classical corpus · contemporary Prophetic Guidance) · source-authority registry, ranked first · school routing · question gate · hard stops and honest limits. Detail in appendix.
The moment we’re in
- Reading a classical Arabic corpus without inventing citations could not have been built before 2024. Each model generation reads it better.
- The category is already monetizing. Usul: $5/$25/$100. Qaf: $19.99/mo ($239.88/yr); $99.99/yr launch offer. Islamify: EUR 5 to EUR 20, no free tier. AI Fiqh: RM 29 to RM 200. All live today.
- Capital is testing this thesis at every scale. Mal raised a $230M pre-launch seed for AI-native Islamic banking (Jan 2026). Bitsmedia, maker of Muslim Pro, raised a $20M Series A earmarked partly for AI in its Quran product (Dec 2023).
- Muslim consumer spending: $2.60T in 2024, projected $3.56T by 2029 (DinarStandard).
Early movers show what to avoid. We build on what they have already learned before the category settles.
Overview: two pools, built bottom-up
Context: about 2 billion Muslims. Fastest-growing major religion, projected to become the largest by around 2075 (Pew). Context only; never our TAM.
English-speaking Muslims willing to pay $15 to $35 a month for cited, school-aware answers
Our own model forecasts 1,347 paying subscribers at the end of year one, from a 60-day launch cohort
Purpose: prove quality and distribution before institutions engage
Research institutes, seminaries, universities with Islamic studies departments, faith-based colleges, continuing-ed academies
Priced on B2B tiers, no client names disclosed
Model’s own evidence: white-label share of revenue 15% to 26% to 34% across three years
Every tool this tradition adopted (print, audio, digital libraries) moved from suspicion to standard. AI under scholarly audit is next. Consumers first to build trust. Institutions follow.
Earned first, paid measured
No CAC benchmark exists for this audience in any public source, so the subscriber forecast assumes no paid lift. We will price the channels with our own launch data.
The same engine, pointed at social.
Tagged into a claim, it replies with the source attached.
It does not argue. Live on X at @ask610; Instagram next. Every reply is a cited answer in front of people who never asked for an ad.
Scholars testing in Canada and London and from the Habibi Tech Summit.
In this tradition credibility is passed hand to hand, not bought.
$9K on Meta video at launch, into our own landing page.
It is there to measure our funnel.
Of the four Muslim consumer products running paid acquisition, three run only on Meta, and none run on Google Search. Every ad observed was a video.
Stated honestly
- Live since March 2026. Shared with Sheikh Abu Eesa Niamatullah ahead of the Makkah and Madinah demo, April 2026
- Two retrieval lanes live: classical corpus, contemporary Prophetic Guidance
- He passed it to scholars in Canada and London for testing. V3 was built from that feedback, shown at the Habibi Tech Summit, and is live at ai.ask610.com
- Informal scholarly advisory: Abu Eesa, all his content approved for inclusion, ongoing conversations
- 168-answer baseline scored across six axes. Scoring governance sealed, auto-fail rules locked
- One product, two audiences: Simple and Scholarly modes together
- Source Authority Registry Tier 1 shipped: 393 of 400 sources placed, gaps documented
- Pipeline: 0 signed, 0 active/qualified, 9 early prospects. Outreach began at the summit by design
JaleesBench baseline: Ask610 ranked #2 of 13 AI systems
| Top 5 Leaderboard | ||
|---|---|---|
| #1 | Ansari | +0.48 |
| #2 | Ask610 | +0.43 |
| #3 | GPT-5.5 | +0.28 |
| #4 | Claude Sonnet 5 | +0.27 |
| #5 | Inkling | +0.26 |
JaleesBench (IASER, Dr. Waleed Kadous et al.) tests whether AI religious guidance holds up under adversarial pressure. Each system answers a question, then the same question is asked again with pushback. If the answer shifts, the system loses marks.
Ask610 was the only system whose score improved under pushback. Every other system’s score dropped. Ask610’s rose by +0.13.
Ask610 evaluated its answers using IASER’s published JaleesBench framework, with external questions, external proof texts and independent LLM judges, scoring #2 of 13 systems tested. An official IASER-administered run is planned for Q1 2027.
A scholar-driven place for vetted Muslim knowledge, in every language, for every age
One trusted source. Many doors. The Expected case assumes these features lift conversion to 7%, a planning assumption.
- Ramadan features: topic pages, 30-day journey, daily āyah, Ḥadīth and duʿāʾ share cards
- Interactive Learning Corner: learn the 99 Names of Allah and improve your duʿāʾ in Ramadan
- iOS and Android apps
- Research Mode for scholars: takhrij, isnad tracing, translation, drafting
- Institutional engines and branded instances on one platform, shared-tenant tier to follow
- A scholar per discipline: Fiqh, Sīra, Aqīdah, Ḥadīth
- Every language: Arabic, Urdu, Turkish, Bahasa, Farsi, Bengali, so the distant library finally opens its doors. Retrieval stays in source languages
- Ask610 for children: its own vertical and AI video, led by its own scholar
- Let truth travel as swiftly as rumor: Ask610 on Instagram and other social platforms, as it is on X today.
Side note: Ask610 ran a self-run JaleesBench evaluation. Results are on the previous slide.
Building for generations: a lasting home for vetted Muslim knowledge
Two engines
Define once: B2B means dedicated app and data with shared retrieval. White Label means fully isolated app, data, and retrieval.
Verified vs hypothesis: institutional service contract precedent exists (Usul). A proven paid consumer model in this category does not exist yet.
| Engine | Offer | Price | Status |
|---|---|---|---|
| B2C subscriptions | Free 1M tokens/mo, Starter 2M tokens, Research 5M tokens | $0, $15/mo, $35/mo | Provisional; free through Ramadan 2027 |
| White Label, Entry (fully isolated) | Own app, data and retrieval, 5,000 answers/mo, $0.40 overage | $4,500/mo (~$54K/yr) + $35,500 setup | Modeled |
| White Label, Standard (fully isolated) | Own app, data and retrieval, 25,000 answers/mo, $0.40 overage | $13,750/mo (~$165K/yr) + $35,500 setup | Modeled |
| B2B pathway, new | Dedicated app and data, shared retrieval, 1,000 answers/mo, for seminaries and smaller organizations | ~$15.9K recurring annual + setup | Modeled separately; not in the base forecast until customer mix and demand are validated |
They charge for features. For us, the source is the product.
Financial plan: expected, floor, growth
| 2027 | 2028 | 2029 | |
|---|---|---|---|
| Revenue, Expected | $325.6K | $506.6K | $778.8K |
| White-label share of revenue | 15% | 26% | 34% |
- 36-month peak cash need: Floor $0.88M · Expected $0.27M · Growth $0.27M
- Expected = operating plan · Floor = capital bound · Growth = execution upside
- Path to self-funding: Expected turns cash positive on a trailing 12-month basis in June 2029; Growth in January 2029.
- AI cost per answer: $0.017, measured on live usage. A $0.35 stress case is kept for capital planning.
- All figures from the pro forma, October 2026.
Expected case: revenue grows 2.4x in three years to $778.8K, and Ask610 turns cash-flow positive in 2029 on this raise alone.
The ask
Sized to the Floor end-2028 requirement ($621.5K), a Q1 2029 cushion ($73.8K), founder contingency ($42.1K) and a $250K growth allocation for marketing, mobile apps and scholar review.
Two horizons: the raise covers the full 36 months on both Floor (peak need $0.88M) and Expected (peak need $0.27M).
A second raise is planned before the end of Year 3 to fund the next stage of growth; it is not required to complete the 36-month Expected plan.
What this raise delivers
What this raise delivers in 12 months: public launch in January 2027, iOS and Android apps, Learning Corner live, first white-label client targeted for late 2027.
A founder’s word
In 2016, in Makkah, I carried a question I could not voice, even to the scholars beside me. Years later, after reaching the peak of my corporate career in New York and then in Silicon Valley, cancer brought my world to a standstill. In that silence, I had to face what remains when everything else is taken away. Faith no longer remained a ritual; it became a place of safety. Knowledge, no longer a luxury, became a necessity. The answers were still very far away, obscured by embarrassment and silence.
I’m not a scholar by any means, but I know what it’s like to go looking for an answer and come up empty. I’ve spent over 20 years running hospitality operations for a range of clients (Google, Palantir, Atlassian, Duolingo, to name a few), where service and efficiency go hand in hand with a bit of ingenuity. When new technology came along, I was always the first to pilot it, shape it, and train our teams to use it. Curiosity, tenacity, and a real desire to build were the things that got me through. When AI started to fundamentally change the world, I built Ask610 myself, using AI as my tool. I broke it, I rebuilt it, I cried many tears, I went to the scholars for advice, and rebuilt it again to meet their standards. When it got beyond my capacity, I brought in an engineering team.
Ask610 is the culmination of my journey. My vision is a sanctuary for the seeker: a private, safe space where questions are met with respect, knowledge is carefully selected and presented, and no one is left to search in isolation. Ask610 is not a chatbot. It’s about putting knowledge back where it belongs: accessible, genuine, and life-changing.
That’s what I’m building. That’s what I needed most.
Who's behind this
| Who | Role | Note |
|---|---|---|
| Sunayna Agha | Founder, Ask610 | Prior corporate operations experience. Owns product, corpus governance, and the financial model. 100% founder-owned; all IP assigned to the entity. |
| Hassan Baig | Engineering lead. CTO of his own firm, contracted with his team | Selected after screening 56 candidates and testing 10 finalists. Built V3, live since the September 2026 summit; now building the January launch. Executed NDA, IP assignment, and corpus agreements. Funded as a full-time role from Year 1 in the model. |
| Sheikh Abu Eesa Niamatullah | Scholarly advisory, informal | Ongoing conversations; full permission to use his content and corpus. Not a contracted partner. |
Board and advisory bench: none formal yet. Conversations underway since the Habibi Tech Summit, September 2026.
Hiring with this raise: senior engineering lead, 1.0 FTE (Y1) · technical support engineer, Pakistan (Y1) · content, 0.25 FTE, 10 hrs/week (Y1) · B2B sales and business development, 0.5 FTE (Y2). Support/community hire triggers in Y3 on measured load. Third engineer and surge cover are conditional, not in the base case.
Your question deserves a real answer
- Ask610: Fourteen centuries of Islamic scholarship, and the tool that carries it forward.
- Scored, scholar-governed answers. Differing views, side by side. Honest about its limits.
- The ask: $1,000,000. Funds the Expected plan through the full 36-month horizon.
- Live app and Ask610 on X. Scored baseline. Scholar advisor. Launching January 2027.
Available on request
- AFull four-school answer: the wudu demo, complete output with citations
- BJaleesBench: self-run baseline October 2026, ranked #2 of 13, most steadfast system (+0.13).
- C168-answer baseline: methodology and results summary
- DSource Authority Registry: Tier 1 structure, the seven documented gaps
- EFinancial model detail: full pro forma, Floor / Expected / Growth, Break-Even Bridge
- FB2B pathway: Tier B scenario record
- GCompetitive captures: Usul, Qaf, Ansari, Islamify, AI Fiqh, Dalail, dated August 25 to September 1, 2026
- HNetwork and early prospects: named relationships and pipeline register
- IArchitecture overview: two retrieval lanes, verification layers
- JLegal structure: entity, IP assignments, corpus agreements