Senior experience is scarce
The marketer who has run hundreds of accounts, the editor who knows a story, the researcher who spots a buyer early. Scarce, costly, and in one place at a time.
Chapter one
Story
The problem and what we do about it
Business BlueprintAmman, JordanOctober 20266 min read
We don’t rent agents. We rent experience.
Senior AI employees and AI-ready teams for companies in Jordan and the Arab world.
Scroll to beginThe problem
Two things hold companies back
Teams don’t work with AI yet
When a team uses AI like a search box, the tools change but the work stays the same. The customer feels no difference.
Companies need experience inside the work, and people who know how to use it.
What we do
We make companies AI-ready
AI employees
Senior experience you hire by role. Trained by experts, working in your private workspace.
Salma · Max · Steve · HusseinAI-ready teams
We train your staff on their own workflows. How they work and think changes, not only their tools.
Programs for companies and schoolsMr. AppDeen
Our orchestrator takes tasks from owners and staff, assigns the right AI employee, or builds a new one.
Build your own agents for freeThe result: better service quality, lower costs, and hours back every week.
Chapter two
Product
How the work gets done
How it works
Mr. AppDeen runs a creative project
- 01
Brief in
“Launch our summer collection on Instagram.”
- 02
Mr. AppDeen plans
Splits the brief into copy, design and ads, and assigns each part.
- 03
The team builds
Each AI employee hands its work to the next. Mr. AppDeen checks it.
- 04
You approve
One tap and it goes live. Every approval trains the team.
New feature
Meet the Roundtable
Several AI models sit at one table and work a question out together.
Discuss
Each model gives its view on the question.
Negotiate
They push back on each other and trade off until one plan holds.
Agree
You get one answer, with the reasons behind it.
AI readiness
Five steps to an AI-ready company
- 01
Map the work
Find where time, money and quality leak.
- 02
Train the people
Hands-on sessions on their own workflows, used the next morning.
- 03
Add AI employees
Senior roles go to trained AI employees, with human checkpoints.
- 04
Measure results
Hours saved, response time, cost per task and quality, before and after.
- 05
Keep improving
Every approval trains the employees. Mr. AppDeen finds the next task.
Training is how we start. AI employees are how we stay.
Chapter three
Edge
Why now, why we win and how we compound
Why now
The market is moving our way
$858M
Capital
MENA AI venture funding in 2025. Pre-seed and seed AI deals rose 56%.
Sept 2026
Compute
Klstrai, Jordan’s first local AI compute platform, opened in Aqaba. Nujum16 is building a 100 MW sovereign AI campus, first phase in 2027.
March 2024
Rules
Jordan’s data protection law came into force. It limits sending personal data abroad, which favours AI hosted in Jordan.
YC 2026
Investors
Y Combinator’s 2026 requests ask for AI-native companies that sell the finished work itself. That is our category.
Sources: Magnitt via Enterprise · Middle East AI News · Nujum16 · Clyde & Co · VC Cafe
Our edge
We win on expert training
Trained by senior experts
Experts write each role’s playbooks, rules and red lines, then grade the work. That becomes the employee’s skills and tests.
Data from this market
What sells in Jordan, which signals come before a purchase. Every client adds to it. Generic agents never see it.
Local depth
Arabic and its dialects, local platforms, local rules and buying habits.
Works on any model
The know-how lives in skills and tests, not in one model. Better models make our employees better. Cheaper models raise our margin.
How we compound
Every client sharpens the craft
Every client’s work makes our AI employees better for all clients.
Private Company Brain
Each client’s AI employee learns from their knowledge, history and approvals. Raw data never leaves their workspace.
Experts review lessons
Only de-identified lessons reach our senior experts, who update the playbooks, rules and tests.
Shared Skills Library
The improved skills ship to every client’s AI employee, one library per role.
No client’s data is shared. Only lessons travel.
Chapter four
Proof
The platform and the clients behind it
The platform
One platform runs seven teams
Mr. AppDeen takes tasks from owners and staff, and hires, builds and trains the AI employees.
Seven teams
Newsroom
Al Rai AI · Hatem AISales
MenaitechGrowth
7 client projectsCommerce
DNA · ToytopiaOperations
Salma, livePulse
TrendJackerEducation
Class AI · Um HatemBlue: the teams we lead with
Shared by every team
Skills Library
Expert-trained know-howCompany Brain
Private per clientAppDeen Credits
Runs every taskModel router
Any model, any hostClients
Clients across four industries
Media
Al Rai
Al Rai AI lets readers chat with the news and hear articles as podcasts. Hatem AI revives Al Rai’s kids’ magazine.
B2B software
Menaitech
An AI lead engine that finds likely HR-system buyers in seven markets. We own it; Menaitech is the launch partner.
Retail
DNA Lifestyle and Toytopia
Shopify build, Merchant Center, SEO, analytics and Google Ads, run with Max, our marketing AI employee.
Education
Stanford-Binet, Inside Brain
SEO, GEO and Google Ads for stanford-binet.org and inside-brain.org, plus AI education programs with Class AI.
Chapter five
Business
How we earn and protect margin
Business model
We charge for experience
Monthly
Pre-trained AI employees
Senior experience, trained by our experts and ready on day one.
Free
Your own agents
Any team can build its own agents with Mr. AppDeen, at no cost.
Prepaid
AppDeen Credits
Every task runs on credits, whether the agent is ours or yours.
Around the platform
AI-readiness programs. Training for companies and schools, per cohort.
Custom builds. The client funds the build and we keep the IP, like our Menaitech lead engine.
Margins
Credits protect our margin
Token prices are public and falling.
Qwen3.6 27B costs $0.30 per million input tokens on OpenRouter. Reselling tokens is a price war.
We sell credits priced on the work.
Each task goes to the cheapest model that passes its test. The spread is our margin.
We self-host only when it pays.
A rented H100 costs about $1,700 a month and beats API prices only at roughly 900 to 3,000 tokens per second, around the clock.
Our path to AI hosted in Jordan
Next
Pilot on Klstrai, Aqaba Digital Hub’s local AI compute platform, live since September 2026.
2027
Become a launch application partner for Nujum16’s first 2 MW phase, targeted for Q1 2027.
Then
AppDeen Sovereign: the same employees, hosted in Jordan, for government, media and banks.
Sources: OpenRouter · Akash (SemiAnalysis data) · Middle East AI News · Nujum16
Chapter six
Plan
Where we are going and the team
The goal
The AI hub of the Middle East.
The platform where the region’s companies build, hire and run their AI workforce.
Always on the newest tech
We test new models, tools and features on real client work and ship what improves results.
Make every client agentic
Train the team, add AI employees, run the work through Mr. AppDeen, and measure.
Free to build, credits to run
Any company can build agents on AppDeen. Every task runs on our credits.
Learn from every task
De-identified outcomes from every client sharpen the Skills Library.
Team
The people behind it
Hamza Abdeen
Director
Let’s do the doctors’ method.
Mahdi Haroun
AI Engineer
I asked ChatGPT, what can’t you do?
Ruqaya Al-Hakkak
Business Developer
Actually?
Osama Al-Hasan
General Manager
It is what it is.
Our work so far
- Salma, live on our platform
- Al Rai AI and Hatem AI, with Al Rai
- AI lead engine with Menaitech, in seven markets
- Shopify and AI-run ads for DNA and Toytopia
- TrendJacker, built on our Arousal Thesis
- AI programs for schools, teachers and companies
We don’t rent agents.
We rent experience.
We are raising our pre-seed round.
appdeen.aihello@appdeen.aiAmman, Jordan