About the role
Ready to apply for roles like this?
Unlock the company name and direct application link. Subscribers get instant access to fresh jobs across Dubai, Abu Dhabi and Riyadh, many with visa support.
Unlock employer & apply directly
The AI Lab builds agentic AI products across the 7X group, from customer service and logistics journeys to internal assistants and government service workflows. Forward Deployed Engineers are how those products reach the business. You embed with an operating team, understand their problem firsthand, and build the solution with them, on their data, inside their constraints.
You own the full loop: frame the business problem, scan what the market has already proven or failed at, scope the smallest version that delivers measurable value, build it, and defend it with evidence.
What you will do
- Embed with business units across the group to identify where AI resolves real operational pain, not where it merely demonstrates well.
- Build working prototypes fast: chat and agentic interfaces, controlled backend actions, verification and eligibility rules, audit trails.
- Work with real operational data, which means messy data. Profile it, decide what to trust, flag what you can't, and never let a system act on records it shouldn't.
- Scope ruthlessly. Define what's in, cut what isn't, and write down why.
- Measure honestly. Label what is measured, what is assumed, and what needs a pilot to prove. No invented numbers.
- Carry validated prototypes toward production: integrations, identity, guardrails, monitoring, handover to engineering and operations.
- Present and defend your work to business owners and leadership.
What we look for
- Strong engineering fundamentals: APIs, data handling, testing, and enough architecture judgment to separate language understanding from business rules and writes.
- Hands-on experience building with LLMs and agent frameworks, including guardrails, evals, and human handoff.
- Product judgment. You can tell the smallest valuable version from a feature list, and you say no with reasons.
- Evidence discipline. You distinguish facts, claims, and assumptions, and you treat vendor numbers as marketing until proven on our workload.
- Care for data protection. You treat customer data as a liability to guard, not a convenience, and you understand what UAE PDPL requires in practice.
- Comfort with ambiguity and bilingual (English and Arabic) contexts is a plus.
How we assess
A practical assignment on a realistic business case, a working prototype, and a 30-minute defense. We evaluate problem framing, scope discipline, prototype quality, and the honesty of your claims. The defense matters as much as the artifact.
