Where we are
Madar AI is pre-revenue. We are in the launch-cohort phase: a small group of MENA subscription-app founders are onboarding to Madar Audit ($299 one-shot) and Madar Forecast ($299 one-shot) as the first paid surface, with Madar Creative ($99/mo) and the full platform following over the rest of 2026.
We are not announcing customer counts, ARR, or LOIs we haven't signed. When we have numbers worth publishing, we will publish them with the methodology beside them.
- Stage. Pre-seed. Bootstrapped to MVP, opening a small round to extend runway and ship the multi-agent system.
- Team. Founder + a small build group operating out of Istanbul under Zyrix Global Technologies.
- Product surface live today. Marketing site (EN/AR/TR), waitlist for the full platform, paid landing for Audit/Forecast/Creative, founder outreach program.
- Product surface in build. Multi-agent growth brain (Oracle, Forge, Sentinel, Curator, Scout, Drafter), benchmarks engine, autonomous-action approval flow.
Team
Madar AI is built in Istanbul by Zyrix Global Technologies. The founder leads product, engineering, and the early go-to-market motion across Türkiye, KSA, UAE, and Iraq — markets the team is going to market in directly rather than through a distributor.
Founder bio, photo, on-the-record quote, and the full team page (including advisors as we add them) live on /about.
The thesis
Three claims that sit underneath the company:
- MENA subscription apps are under-tooled. AppsFlyer, Adjust, RevenueCat and the rest were built for English-language US/EU growth teams. Arabic creative measurement, iyzico/local-IAP economics, Ramadan-distorted cohorts, and PDPL-sensitive data handling are second-class citizens in every existing stack.
- The next layer is agentic, not dashboard. Founders do not need another chart wall. They need a system that ingests the same data, decides what to do, and executes — with human approval at the right gate. The unit of value shifts from "surfaces metrics" to "closes a loop."
- Region-native wins this category. The team that ships in Arabic first, prices in local currencies first, and signs DPAs in the right jurisdictions first will be the default growth layer for the MENA app ecosystem before US incumbents notice it is a category.
Market frame
We are deliberately conservative on TAM language. The honest framing:
- Direct serviceable surface. Subscription mobile apps in MENA (KSA, UAE, Egypt, Türkiye, Jordan, Morocco, Kuwait, Qatar) doing meaningful spend on UA — meditation, fitness, edtech, finance, language-learning, productivity. Several thousand apps, a few hundred of them with budgets that justify our entry price point.
- Adjacent surface. App studios and growth agencies that manage portfolios of subscription apps for the region. One agency contract is often worth a dozen single-app contracts.
- What we are not claiming. We are not claiming TAM numbers pulled from third-party analyst reports we have not verified. If a slide in a deck quotes a TAM, the source is named on the same slide.
Product shape
Madar is a multi-agent system, not a dashboard. Six specialized agents share a working memory and a planning loop:
- Oracle — forecasts revenue, retention, and LTV; flags trajectory deviations.
- Forge — generates and tests Arabic, Turkish, and English ad creative; closes the loop into ad-platform APIs.
- Sentinel — watches paid acquisition spend, integrations health, and fraud signals; pauses when needed.
- Curator — assembles weekly and monthly briefings tailored to the founder's actual portfolio shape.
- Scout — pulls benchmarks from the cohort and flags where the app under-indexes against peers.
- Drafter — drafts the next action (campaign edit, creative rotation, paywall test, cohort dig) and queues it for one-click human approval.
The platform commits to a single rule: every action that touches a third-party system requires a human approval gate. "Autonomous" is what we mean by the agents proposing the action — not by them executing without consent.
Why we win
- Regional data flywheel. Each onboarded MENA app strengthens the benchmark engine for the next one. US incumbents do not have this corpus and would have to compete with us in a market that does not look like their home market.
- Arabic-first creative loop. Forge is built on Arabic dialect handling and right-to-left UX from day one. Retro-fitting this onto an English-first product is a multi-quarter rewrite.
- Compliance posture as a sales tool. PDPL, KVKK, and GDPR-readiness from day one lets us sell into KSA and Türkiye finance, health, and edtech surfaces that US-only competitors cannot legally serve.
- Price discipline. We open with a $299 one-shot Audit and a $99/mo Creative tier. This puts Madar inside the budget of a founder running a single subscription app — a segment most of the incumbent stack is already too expensive to serve.
Use of funds
If you are reading this as part of due diligence: the round is structured to fund 18 months of build and go-to-market with a clear bar for the next milestone. The allocation:
- Engineering. Land the full multi-agent system, the autonomous-action approval flow, and the benchmarks engine at production quality. This is the majority of the round.
- Cohort onboarding. Hands-on onboarding for the first 50 MENA apps, including bespoke integration work where AppsFlyer / iyzico / local-IAP setups are non-standard.
- Compliance. SOC 2 Type II, PDPL filings in KSA and UAE, KVKK posture documentation in Türkiye, and the legal work to sign DPAs in each jurisdiction without external counsel as a bottleneck.
- Go-to-market. Founder-led outreach across MENA, a small content surface (blog, case studies, benchmarks reports), and the events / dinners format that has historically converted in this region.
We are not buying ads at this stage. The first 50 customers come from founder-to-founder relationships, not paid acquisition.
What could break this
An investor who has done diligence wants to know the risks before being told the upside. The honest list:
- Multi-agent reliability. Agentic systems are still a young category. A meaningful share of engineering effort goes into making the agents safe to hand a credit-card-spending decision to, not into the agent loop itself.
- Distribution lag. Founder-to-founder selling in MENA is slow. We are explicit that the first 12 months are go-to-market-cost-heavy and revenue-light.
- Incumbent reaction. AppsFlyer, Adjust, or RevenueCat could ship an Arabic-first or MENA-specific feature set. The bet is that doing so requires them to rebuild a regional sales motion, not just a feature.
- Talent. Hiring Arabic-fluent growth-engineering talent in Istanbul is harder than hiring generalist engineering talent. We are explicit about this in our pipeline planning.
Request the deck
Investors only. We send the current deck within 24 hours alongside the next-step options (call, written follow-up, NDA pack). One sentence on your fund's focus is enough — we will calibrate the rest from there.
Talk to us
If the thesis lands and you want to spend an hour going deeper — financial model, agent demos under NDA, customer pipeline, references — email investors@madarai.co with one sentence on your fund's focus and we will respond within 48 hours.
We are not currently using a placement agent. All conversations are direct with the founder.
