Mid-Senior AI Engineer - Generative AI & Agents

• Bachelor's degree in Computer Science, Software Engineering, Artificial Intelligence, Data Science, or a related technical field. • 4+ years of software engineering experience, including approximately 1-2+ years of hands-on experience building LLM or Generative AI applications. • Strong programming skills in Python. • Strong understanding of software engineering principles, APIs, distributed systems, and production application development. • Practical experience building RAG applications, including embeddings, vector databases, document processing, and retrieval techniques. • Experience designing and developing AI agents, tool-calling systems, and multi-step LLM workflows. • Hands-on experience with AI frameworks such as LangChain, LangGraph, Semantic Kernel, LlamaIndex, or equivalent. • Hands-on experience with major LLM platforms such as Claude, OpenAI, Azure OpenAI, Gemini, or equivalent. • Experience integrating AI capabilities into production applications and backend services. • Understanding of AI evaluation, prompt testing, RAG evaluation, and AI application quality measurement. • Understanding of production AI concerns including reliability, latency, scalability, security, observability, and cost. • Strong product mindset with the ability to translate AI capabilities into practical business solutions. • Strong analytical and problem-solving capabilities. • Good communication skills and ability to collaborate across Product, Engineering, Data, Security, Platform, and Operations teams

About Madar Madar is a leading Saudi digital logistics platform, owned by Elm, transforming transportation management across industries. Madar connects shippers, carriers, and logistics partners through a unified ecosystem that improves visibility, efficiency, automation, and shipment execution. The platform enables seamless shipment management, financial integration, real-time tracking, and operational visibility, helping businesses move freight with greater speed, transparency, and confidence. About the Role Madar is looking for an experienced and hands-on Mid-Senior AI Engineer to design, build, and productionize AI capabilities across our logistics platform. The role combines Generative AI, AI agents, Retrieval-Augmented Generation (RAG), intelligent automation, and applied machine learning to solve real logistics and operational problems. You will work closely with Product, Engineering, DevOps/Platform, Data, Security, and Operations teams to take AI solutions from business requirements and prototypes into reliable production services. The role will also help drive AI adoption across Madar's engineering organization by establishing reusable AI development patterns, tools, standards, and engineering practices. The ideal candidate combines strong software engineering fundamentals with hands-on experience building production-grade AI and LLM applications. Key Responsibilities AI Product Development • Design and build AI-powered logistics capabilities such as shipment assistants, customer-support assistants, operational copilots, document intelligence, exception-handling workflows, and intelligent automation. • Contribute to predictive AI use cases such as ETA prediction, anomaly detection, operational risk identification, and shipment-related forecasting where appropriate. • Translate product, business, and operational requirements into practical and reliable AI solutions. • Work closely with Product and Engineering teams to move AI capabilities from proof-of-concept to production. • Design reusable AI components and services that can be consumed across multiple Madar products and engineering teams. • Integrate AI services with Madar's existing Angular, Node.js, API, event-driven, and backend platforms LLM & RAG Engineering • Design, build, and maintain Retrieval-Augmented Generation pipelines using embeddings, vector databases, structured data, unstructured documents, and logistics-domain information. • Implement effective document ingestion, chunking, indexing, retrieval, reranking, and context-management strategies. • Build AI applications using leading LLM platforms such as Claude, OpenAI, Azure OpenAI, Gemini, or equivalent enterprise AI platforms. • Use AI frameworks such as LangChain, LangGraph, Semantic Kernel, LlamaIndex, or equivalent frameworks where appropriate. • Design model-agnostic architectures that allow Madar to evaluate and adopt different models based on capability, reliability, latency, security, and cost. • Improve retrieval relevance, grounding, response accuracy, and overall application reliability. • Design structured output and tool-calling patterns to reliably connect LLMs with business applications and backend services. AI Agents & Intelligent Automation • Design and develop AI agents for logistics workflows such as dispatch assistance, customer support, shipment exception handling, operations support, and internal productivity. • Build reliable agentic workflows that combine LLM reasoning, deterministic business rules, APIs, enterprise data, and external systems. • Implement appropriate state management, memory, tool usage, workflow orchestration, and human approval steps. • Define boundaries between AI-driven decisions and deterministic application logic. • Design human-in-the-loop workflows where AI-generated decisions or actions require review or approval. • Ensure AI agents operate within clearly defined permissions and business boundaries. AI Evaluation & Quality Engineering • Design evaluation frameworks and test datasets for measuring AI system performance. • Evaluate solutions across dimensions including retrieval quality, groundedness, hallucination rate, task completion and success rate, response quality, accuracy, latency, reliability, token consumption, and cost. • Establish regression testing for prompts, retrieval pipelines, models, tools, and agent workflows. • Develop golden datasets and benchmark scenarios for critical Madar AI use cases. • Apply human evaluation, automated evaluation, and LLM-as-a-judge approaches where appropriate. • Continuously measure production AI performance and identify opportunities for improvement. AI Security, Governance & Responsible AI • Apply secure AI engineering practices to protect Madar's customers, systems, and business data. • Implement controls against prompt injection, indirect prompt injection, sensitive information leakage, unauthorized data access, excessive agent permissions, unsafe tool execution, and cross-user or cross-tenant data exposure. • Design secure authorization models for RAG applications and AI agents. • Ensure AI applications respect existing application-level access controls and data boundaries. • Apply secure secrets management and authentication when integrating AI services with internal and external systems. • Support auditability by implementing appropriate logging, tracing, and monitoring of AI interactions and actions. • Work with Security and Platform teams to establish practical AI governance and security standards Engineering AI Adoption • Drive practical adoption of Claude Code and other AI-assisted development tools across Madar engineering teams. • Develop reusable AI skills, subagents, prompt templates, engineering instructions, and development workflows. • Help engineers use AI effectively for coding, unit and integration testing, code reviews, debugging, documentation, refactoring, applic