Swibit is an AI solutions company operating in the MENA region, the UK, and the US. We specialise in delivering high-end consulting and building scalable digital platforms, AI-powered products, and strategic solutions across sectors such as government, public safety, real estate, logistics, education, and public services.
Our model is global design and local execution — combining international expertise with deep local market understanding to deliver impactful, future-ready solutions aligned with national digital transformation visions.
Swibit is actively expanding in Oman, working on strategic engagements with government and enterprise clients in areas including AI consultancy, digital transformation, and technology modernisation.
Role OverviewWe are looking for an AI Engineer based in Oman to build, fine-tune, and deploy AI/ML models and intelligent systems for Swibit's client engagements in the Omani market. You will work on production AI systems — from Arabic NLP and document intelligence to recommendation engines, agentic AI workflows, and LLM-powered applications — serving government and enterprise clients.
This is a client-facing technical role. You will work on-site in Oman, collaborating closely with Solution Architects, Business Analysts, and Project Managers on delivery, while working day-to-day with Swibit's AI team across Iraq and Pakistan on engineering execution.
Key ResponsibilitiesAI/ML Development
Build, train, fine-tune, and evaluate machine learning models for production use
Develop NLP solutions with a focus on Arabic language — text classification, NER, summarisation, sentiment analysis
Build and maintain LLM-powered applications — RAG pipelines, agentic workflows, prompt engineering
Implement recommendation systems, search, and ranking algorithms
Work with computer vision models where required (object detection, image classification)
Production & IntegrationDeploy models as production-grade APIs using FastAPI and containerised services
Integrate AI capabilities into client platforms and Swibit's backend services
Optimise model performance — latency, throughput, cost, and accuracy trade-offs
Build evaluation pipelines and monitoring for model performance in production
Manage data pipelines for training, validation, and inference
Ensure AI systems respect data sovereignty and client privacy requirements — critical for Omani government engagements
Client & Delivery SupportSupport Solution Architects in designing AI components of client solutions
Participate in client workshops and technical discussions alongside the delivery team
Translate client business problems into AI/ML solutions
Support pre-sales with AI feasibility assessments and proof-of-concept development
Collaborate with Backend Engineers to integrate AI into product workflows
Stay current with the AI/ML landscape — new models, techniques, and tools
Contribute to Swibit's AI Collective — internal knowledge sharing and best practices
Participate in sprint planning, daily standups, and retrospectives
Mentor junior engineers and support knowledge transfer to client teams
RequirementsBased in Oman or willing to relocate to Muscat
2–4 years of professional experience in AI/ML engineering (not just academic/research)
Strong proficiency in Python — production-grade, clean, well-structured code
Hands‑on experience with Py Torch or Tensor Flow
Experience with
NLP — particularly transformer‑based models (BERT, GPT, T5 family)
Experience building applications with LLMs — OpenAI API, Hugging Face, or open‑source models
Familiarity with Lang Chain
, Llama Index
, or similar LLM orchestration frameworks
Understanding of ML fundamentals — training, evaluation, overfitting, feature engineering
Experience deploying models as APIs (FastAPI, Flask)
Comfortable with Git, pull requests, and collaborative development workflows
Strong problem‑solving skills and analytical thinking
Fluency in English (spoken and written)
PreferredExperience with
Arabic NLP — tokenisation, morphological analysis, Arabic‑specific models (AraBERT, Jais, ALLaM)
Experience with RAG (Retrieval‑Augmented Generation) pipelines and vector databases (Pinecone, Weaviate, Qdrant)
Familiarity with…
(If this job is in fact in your jurisdiction, then you may be using a Proxy or VPN to access this site, and to progress further, you should change your connectivity to another mobile device or PC).