How we work · GCC
Domain expertise first.
AI second.
We stay until it's working.
A short read on how we engage with founders and operators across the Gulf — the methodology, the pricing philosophy, and the things we won't do.
The thesis
Most "AI consulting" you'll see in the region is the same software, translated, rebadged, and re-sold. We do the opposite. We start by understanding the business — the family-ownership structure, the payment mix, the seasonal rhythms of Ramadan and Hajj — and we treat AI as the execution arm of that domain understanding, not as the product itself.
It's the harder way. It's also why our work compounds rather than decays.
Five pillars we build every engagement on
1. Domain experts shape the system
Before we configure any agent, we sit with the people who actually do the work — owners, managers, front-of-house. The system mirrors their judgement, doesn't try to replace it.
2. Bilingual where it matters
Customer-facing agents handle Arabic and English with native fluency. Internal dashboards, reports, and admin can be in whichever language your team prefers.
3. Honest about what doesn't work
Three engagements out of ten, our recommendation is "you don't need us yet — fix this first, then come back." We say it on the first call. It's the one filter that keeps us aligned with the clients we serve.
4. Pricing is a fraction of value
If the system can't pay for itself in 90 days at a 3-4x ratio, we don't take the engagement. The math has to work. Pricing in SAR, transparent, no long-term contracts.
5. We stay while it compounds
Most AI consultancies disappear after launch. We're built differently — every engagement includes the human who keeps the system tuned as your business changes. That's how AI keeps paying its bill in month 6, not just month 1.
What an engagement actually looks like
- Week 0 — Discovery. Free 45-minute call with Saif. We understand the business, identify the leak, decide whether AI is the right shape for it. About 30% of the time we say "no, here's what to do instead" — that's healthy.
- Weeks 1-2 — Scope & price. We send a written 2-page plan with the specific agents we'd build, the integrations, the math (recoverable revenue vs. monthly fee), and a SAR price. You either accept, push back, or pass — no pressure.
- Weeks 3-5 — Build. We wire integrations, define agent tone in your language, load your domain knowledge. Weekly check-ins with your team.
- Weeks 5-8 — Tune. Go-live, then four weeks of active tuning — call recordings sampled, prompts adjusted, edge cases captured. This is the messy part where most "AI launches" silently fail. We're there for it.
- Month 3+ — Review & evolve. Monthly call, monthly digest. Adjust as your business changes. We expand the system when it's earning, scale back if something's off.
What we don't do
- Free trials. Free trials select for tyre-kickers. Our system is for paying clients only.
- Replace judgement-bound work with automation. If a person's role is reading tone and building relationships, AI is at most a triage layer — not a replacement.
- Lock-in contracts. Cancel month-to-month after the first 60 days. We earn renewal monthly.
- Hide that it's AI. Customers always find out. We default to identifying agents as automated up front.
Want to know if your situation is the right shape for AI?
Send us one detail about where the friction sits. Saif reads every regional message and we'll send back a 2-paragraph plan — usually within a day. No call required, but if you'd like to meet in Riyadh once the office opens in April, that works too.
Send a message