Senior AI Forward Deployed Engineer

Dicetek LLC; Dubai; On-site; AED 360,000 - 600,000; Full time

Job summary

Dicetek LLC seeks a senior, hands-on AI engineer to ship production-ready AI solutions within live terminal operations contexts. You will embed with operational teams, lead cross-functional squads, and own security, data governance for LLM and agentic systems.

You will stand up MLOps/AIOps foundations, drive codified delivery methodologies, and develop reusable accelerators to shorten time to value for future projects.

Qualifications

• Senior, hands-on engineer with recent production AI delivery experience.
• Production LLM and agentic AI including RAG, orchestration and multi-step workflows.
• MLOps depth: CI/CD, model registries, monitoring and evaluation pipelines.
• Strong Python and modern software‑engineering practice.
• Consulting-adjacent skills to work with non-technical stakeholders and manage end-to-end delivery.
• Self-driven with minimal direction: scope work, set plan and drive to production.
• Technical leadership without formal authority across engineers and staff.
• Strong product mindset: feedback, prioritization, trade-offs, adoption.
• Deep AI evaluation expertise: groundedness, hallucination detection, KPIs, A/B testing.
• Production AI cost optimization: model routing, prompting, caching, cost management.

Responsibilities

• Embed with operational teams to scope, build and ship AI solutions against live business problems.
• Lead cross-functional delivery squads drawn from business and IT without formal authority.
• Take LLM and agentic systems from prototype to embedded production with security and data governance ownership.
• Build accelerators: reusable components, prompts, templates and libraries to shorten delivery time.
• Stand up MLOps/AIOps foundations including CI/CD, evaluation pipelines and deployment standards.
• Codify how AI gets built into a repeatable delivery methodology.

Skills

Production AI

LLM & agentic AI

MLOps (CI/CD)

Python

Stakeholder management

Self-driven

Technical leadership without authority

Product mindset

AI evaluation

AI cost optimization

Domain-agnostic delivery

Job description

Technical Competencies (technical Skills Required To Perform The Role)

Must-Haves

• Senior, hands‑on engineer who ships, with recent production AI delivery experience.
• Production LLM and agentic AI: RAG, orchestration and multi-step agent workflows.
• MLOps depth: CI/CD, model registries, monitoring and evaluation pipelines.
• Strong Python and modern software‑engineering practice.
• Consulting‑adjacent skills to work with non‑technical operational stakeholders, identify the real problem and manage delivery end to end.
• Self‑driven with minimal direction: scope the work, set the plan and drive it to production without waiting for a brief.
• Technical leadership without formal authority across engineers, analysts and operational staff; ability to set adopted standards and influence senior stakeholders.
• Strong product mindset: user feedback, feature prioritization, technical trade‑offs and adoption.
• Deep AI evaluation expertise: groundedness, hallucination detection, task success, latency, safety, business KPIs, A/B testing and continuous regression testing.
• Production AI cost optimization: model selection/routing, prompt and token optimization, caching and inference‑cost management.

Nice‑to‑Have

• Domain exposure to container terminals, maritime logistics, free zones or freight operations, including TOS data, gate and yard processes, customs/trade documentation, or warehouse/contract-logistics workflows.
• GCC or comparable multi‑country regional experience; Arabic is a plus.
• Platform engineering or developer‑experience background, including internal tooling adopted by other engineers.

Principal Responsibilities

• Embed with operational teams — terminal operations, gate and yard planning, trade documentation, logistics and economic zones — to scope, build and ship AI solutions against live business problems.
• Lead cross‑functional delivery squads drawn from the business and IT without formal authority over team members.
• Take LLM and agentic systems from prototype to embedded production, owning security, data governance and operational handover.
• Build accelerators: reusable components, prompt and agent templates, reference architectures and internal libraries that reduce delivery time for subsequent projects.
• Stand up the Lab’s MLOps/AIOps foundation, including CI/CD for models and agents, evaluation pipelines, monitoring, versioning and deployment standards.
• Codify how AI gets built: convert one‑off builds into a documented, repeatable delivery methodology the wider organization can run.