AI Engineer

This is a remote position.

CoorB is looking for a Senior AI Engineer to design, build and deploy AI solutions for digital banking. You will build production-grade LLM applications, including retrieval-augmented systems, knowledge graphs and fine-tuned models, that deliver measurable business value to our banking clients across the GCC.

Key Responsibilities

• Design and build RAG pipelines: document ingestion, chunking, embeddings, hybrid search and re-ranking.

• Build and maintain knowledge bases and knowledge graphs (for example Neo4j, or GraphRAG approaches).

• Fine-tune LLMs (LoRA/QLoRA, PEFT) on domain-specific data, and evaluate them against base models.

• Develop agentic AI workflows, including tool calling, multi-step orchestration and memory.

• Set up LLM evaluation frameworks to measure accuracy, hallucination rate, latency and cost.

• Deploy and operate models in production (LLMOps/MLOps), including monitoring, versioning and CI/CD.

• Apply security, data privacy and responsible AI practices suited to regulated banking environments.

• Work with architects, developers and client stakeholders to turn business needs into AI solutions.

Requirements

Mandatory Requirements

All of the following are required:

• 5–7 years in software, ML or AI engineering, with at least 2 years building LLM-based applications.

• Strong Python; hands-on with LangChain, LlamaIndex or similar frameworks.

• Proven experience building RAG systems in production.

• Hands-on with vector databases (Pinecone, Weaviate, Milvus, pgvector, Azure AI Search or similar).

• Experience with knowledge graphs and graph databases.

• Proven experience fine-tuning open-source LLMs (Llama, Mistral, Qwen or similar) using Hugging Face tooling.

• Experience with commercial LLM APIs (OpenAI/Azure OpenAI, Anthropic, Google Gemini).

• Experience building agentic AI and multi-agent solutions, including MCP (Model Context Protocol).

• On premises, Cloud experience on Azure, AWS or GCP, plus Docker and Kubernetes.

• Solid grasp of prompt engineering, embeddings and LLM evaluation methods.

• Arabic NLP or multilingual model experience.