Job Description

Roles & Responsibilities

Roles & Responsibilities (generic):

• Design, build, and maintain AI pipelines and automations end-to-end (data → model → serving → monitoring).

• Integrate and orchestrate LLMs and generative models via APIs — Anthropic, OpenAI, DeepSeek, and others — plus open-source models.

• Fine-tune / adapt models (LoRA and full fine-tuning) — dataset prep, training, evaluation, and serving.

• Deploy and operate services in production — servers/GPUs, containers, CI/CD, monitoring, and safe rollbacks.

• Build backend logic (Python), basic frontend, and work with databases.

• Leverage AI coding agents (Claude Code, OpenAI Codex, Cursor, Cline, and similar) to ship fast.

• Continuously research and adopt the newest AI tools, models, and techniques.

Desired Candidate Profile

Requirements (must-have):

• Fine-tuning experience — hands-on LoRA / fine-tuning of LLMs or vision/diffusion models (data → train → evaluate → serve). Mandatory.

• Deployment / MLOps experience — shipping to production on Linux/GPU servers (Docker, systemd/CI, git), with monitoring and rollback discipline. Mandatory.

• Strong programming (Python primary) + backend fundamentals; basic frontend (HTML/JS, React or similar).

• Multi-provider LLM API integration — Anthropic, OpenAI, DeepSeek, etc. (API keys, structured JSON outputs, token/cost management, retries/error handling).

• Open-source model knowledge — running/serving models locally (e.g., LLaMA, Mistral, Qwen, Stable Diffusion/FLUX), quantization/inference basics.

• Databases / SQL (Postgres or similar).

• Proficiency with AI coding agents — Claude Code, Codex, Cursor, Cline, and other AI dev tools — used effectively for real engineering work.

• Strong research mentality — stays current with the daily, fast-moving AI landscape and can independently evaluate and adopt new tools/models.

• Great problem-solving, attention to detail, and production discipline.

Nice-to-have (bonus):

• ComfyUI / diffusion workflows, RAG & vector databases, agent/tool-use frameworks, prompt-engineering depth, media/ffmpeg pipelines, cloud/GPU infrastructure, distributed-systems patterns (queues, idempotency,

checkpointing).

Working style (it matters for live production):

• Ownership & autonomy — takes a task, researches, plans, executes, and documents it end-to-e

• Production rigor — verifies before calling anything "done," always backs up + has a rollbace.

• Clear communicator — explains technical decisions in plain English to non-engineers.

• Fast but careful — moves quickly without cutting corners on safety.

Location: Business Bay, Dubai, UAE.

Job Type: Full-time, On-site (no remote option).

Working Hours: 12:00 PM - 9:00 PM, Monday to Saturday. Because our clients and customers are based primarily in the UK, Europe, and the Americas, our working hours are aligned to those time zones to ensure we're available when our markets are active.

Salary: AED 4,000 - 6,000 per month + AED 350 travel allowance + Visa + Insurance.

Experience Required: 3–5 years.