We help companies transform how their teams work with AI, build the agents that do the work, and comply with AI standards. Then we prove the value, use by use.
The Twin: your business at the core, with Transform and Comply in orbit around what we Build.
One team that advises and builds. We start with your people and your numbers, and we stay until the AI is used every week.
We set up your AI Center of Excellence and get every team using AI, measured week by week.
We build the AI that does real work, from first proof to production, with a human check where it matters.
We make your AI auditable and ready for ISO/IEC 42001, without slowing your teams down.
Our engineers and AI Solution Architects join your team, ship with your people and leave the skills behind.
TBC no longer means “to be confirmed”.
It means Transform · Build · Comply: AI your teams use every week, and numbers your board can trust.
If you cannot measure it, you cannot control it.
We track the same eight measures in every program, so leaders see real progress, not slides.
of people using AI each week by quarter end
AI time savings logged in one quarter (≈ 900 person-weeks)
live AI use cases, ranked by Pain × Gain × AI fit
of a team’s weighted work covered by AI it really uses
faster on core engineering tasks, team-reported
each use case graded from chat help to an agent that owns the routine
return on build cost, tracked weekly as Savings ÷ Investment
AI products we own and run: Qarina.ai and detumble.ai
How we count. One tap in our adoption tool = one AI use. The hours per use are set once per use case by an AI committee, so numbers add up the same way in every team. Guards catch back-dating, double counting and over-reporting.
From our 2026 programs · team-reported, logged per use · client names withheld.
Three tools we use in every program. Open any card to see how it works.
Every AI use is counted once, with the hours fixed per use case. So savings and ROI add up the same way in every team.
An AI committee sets the hours per use once. Users only tap. Guards catch back-dating, double counting and over-reporting, and every number can be traced to the use that made it.
A map of every activity a team does, weighted by its share of the week. It shows where AI helps today, and where it should help next.
Twin coverage is the share of a team’s weighted work that AI really covers. We refresh each Twin every six months, so the plan follows how the work changes.
An AI management system that auditors recognise, built into the way your teams already work.
Deep work in a few industries, not shallow work in many.
Regression-log triage, spec and databook Q&A, RTL coding, warning clean-up, place-and-route tuning, test generation.
Talk to usWith detumble.ai: drawings to quantities, Bill of Quantities, material take-off, 3D model from 2D drawings, tender questions.
Talk to usWith Qarina.ai: Arabic legal research with evidence, case reasoning step by step, drafting and review.
Talk to usHR CV screening and headhunting, finance reconciliation and reporting, security and compliance tracking, quality index, CEO reports.
Talk to us Team backgroundSafety-critical software programs and the engineering standards that come with them.
Talk to us Team backgroundResearch in data-driven safety verification for robots and swarms.
Talk to usArabic legal AI. Deep legal reasoning grounded in Egyptian and Arab law and cases. Every answer is backed by evidence.
Construction AI. Drawings in; quantities, Bill of Quantities, material take-off and a 3D model out.
Six ways to start. Each one ends with something your teams keep.
Map every team's work, find and rank AI use cases, and get a roadmap.
Includes: Team Digital Twin · Ranked use-case backlog · Roadmap
Your AI Center of Excellence, set up and run with you.
Includes: Adoption tool and dashboards · AI Transformers network · Education and certification · ROI engine
One workflow from proof-of-concept to production, with tests and human checks.
Includes: Working agent or harness · Tests and human checks · Run book and hand-over
Gap check against 31 base practices and 9 lifecycle gates.
Includes: Gap report · Governance passport per use case · Action plan
Our engineers and architects embedded in your team.
Includes: Monthly package · Price per outcome, not per hour · Hand-over built in
“Mastering Harness Engineering”: a session for leaders and builders.
Includes: How agents, skills and harnesses work · Where AI fits your company · Next steps for your teams
We pick the stack that fits your data, budget and rules: cloud or on-premise.
A small senior team with a wide partner network, based in Cairo and working across Egypt, the Gulf and beyond.
16+ years leading AI transformation and engineering programs across automotive, aerospace, semiconductors and software. PhD candidate in data-driven safety verification for robotics. 9 publications. BSc and MSc in Aerospace Engineering, Cairo University.
Builds with clients across our agents, harnesses and products.
Specialists we embed in client teams to build, ship and hand over.
Send the form and we reply within two working days with next steps.