SENIOR SPECIALIST – AI SOLUTIONS & OPTIMIZATION

Drive the research, design, and deployment of agentic AI systems built on meta-agent architectures and advanced Reinforcement Learning (RL) to deliver breakthrough optimizations across multi container-terminal end to end logistics operations. The role demands strategic vision, deep technical mastery, and an “impossible-is-nothing” mindset to translate cutting-edge research into real-time, large-scale solutions that redefine efficiency, resilience, and profitability while advancing the organization’s digital-transformation roadmap.

Key Responsibilities

• Architect meta-agent frameworks that coordinate multiple autonomous agents for yard, vessel, and gate optimization in real time.
• Design, train, and fine-tune RL / RLHF / RLVR models for dynamic scheduling, resource allocation, and route planning under high-variance, stochastic conditions.
• Build end-to-end ML pipelines—from streaming data ingestion through automated reward-signal generation to continuous policy deployment.
• Benchmark & blend mathematical solvers (Pyomo, OR-Tools, SCIP, COIN-OR) with RL policies to guarantee state-of-the-art performance and stability.
• Research emergent AI techniques (generative agents, quantum-inspired optimizers, MARL) and run rapid prototypes to validate operational impact.
• Lead large-scale data-engineering initiatives (Spark, Hadoop, Kafka, streaming lakes) ensuring low-latency observability for agent policies.
• Deploy & orchestrate solutions on Kubernetes-backed on-prem clusters and multi-cloud stacks (AWS, GCP) with bullet-proof MLOps, CI/CD, and observability.
• Enforce robust AI governance—versioning, bias audits, explainability, and safety constraints across all agentic interactions.
• Mentor & upskill cross-disciplinary teams in RL theory, safe policy evaluation, and reward-shaping best practices.
• Quantify ROI & KPIs (throughput, turnaround time, fuel/energy consumption) and communicate wins to executive stakeholders.
• Reverse-engineer and enhance legacy ML models, embedding agentic components for continuous performance gains.
• Use lean principles to streamline codebases, eliminate technical debt, and accelerate experimentation cycles.
• Integrate LLM-driven NLP modules (chatbots, doc-automation, maritime text analytics) into broader optimization workflows.
• Contribute technical foresight to the corporate digital-transformation roadmap, ensuring long-term scalability of AI initiatives.

Technical Skills

• The ideal candidate commands an end-to-end Python-centric AI stack—mastering TensorFlow, PyTorch, JAX, Ray RLlib, and Scikit-learn and is equally fluent in open-source optimisation libraries such as Pyomo, OR-Tools, SCIP, and COIN-OR to solve complex scheduling, routing, and resource-allocation problems.
• They build and maintain large-scale data-engineering backbones with Apache Spark, Hadoop, Kafka, and Delta Lake, enabling high-throughput ingestion of both structured and unstructured data for real-time analytics that feed agentic RL systems.
• Deployment expertise spans on-prem and multi-cloud environments—Docker, Kubernetes, Helm, Git/GitHub Actions, Terraform, AWS (EKS, SageMaker), and GCP (GKE, Vertex AI)—with disciplined CI/CD, security hardening, and performance tuning.
• End-to-end observability is delivered through Prometheus, Grafana, and the ELK/OpenSearch stack. On the NLP front, the candidate wields SpaCy, NLTK, and Hugging Face Transformers to craft generative-AI chatbots, document-automation workflows, and maritime text-analytics pipelines.