About this role
Job Description OVERVIEW Job Title Senior Engineer Job Code 476244 Grade I2 Group - Division Research & AI Department Applied Research Unit Delivery ROLE PURPOSE The aim is to state the overall significance of the job from the organization’s perspective. The role exists to design, build, evaluate, and deploy applied AI solutions, including LLM-based workflows, vision-language models, and autonomous agents, that turn business use cases into reliable production capabilities. It selects the simplest effective design, grounds every decision in rigorous evaluation against real success criteria, and ensures solutions meet latency, cost, safety, and governance requirements. The role collaborates with Product, Engineering, Infrastructure, and ODC teams to deliver secure, scalable, and policy-compliant AI solutions. KEY ACCOUNTABILITIES & ACTIVITIES This section describes the principal outputs required from the job. Key Accountabilities Key Activities AI Use Case Scoping & Prototyping Translate business use cases into well-defined AI tasks with clear success metrics, selecting the simplest design that meets the required bar, whether a single model call, a fixed workflow, or an autonomous agent. Build prototypes using LLMs and vision-language models to validate technical feasibility on real data. Agent Architecture & Tooling Design agent architectures and the tools they use, including MCP servers, with clear interfaces, descriptions, and error messages that models can use reliably. Build memory and state-handling mechanisms, and determine when to split work across sub-agents, so that agents remain on track across long, multi-step tasks. Context Engineering & Model Adaptation Engineer what goes into the model's context, including retrieval, prompt structure, and compaction strategies for long-running tasks. Fine-tune and adapt models (SFT, LoRA, task-specific heads) where prompting and retrieval fall short, and justify the associated cost with evaluation results. AI Evaluation & Benchmarking Build evaluation datasets and automated evaluation harnesses from the outset of every project. Define metrics tied to real success criteria, such as accuracy, error severity, latency, and cost, rather than relying solely on public benchmarks. Evaluate agents on full trajectories as well as final answers, and design quantitative benchmarks to compare harness configurations on large, multi-step tasks. Error Analysis & Continuous Improvement Conduct error analysis to identify failure modes, and drive systematic improvements to models, prompts, and harnesses. Share findings, reusable components, and known model gaps with the product team to inform solution and roadmap decisions. Inference & Production Deployment Define and build inference pipelines, selecting the serving setup, batching, quantization, and caching approaches required to meet latency and cost targets. Deploy models and agents to production, and monitor quality, drift, cost, and step-by-step agent behavior once live. AI Safety, Guardrails & Access Control Implement guardrails against hallucination, prompt injection, and unsafe outputs, and document known limitations. Scope agent permissions and sandboxing, and enforce human approval for high-impact actions. Policies, Processes & Procedures Follow all relevant departmental policies, processes, standard operating procedures, and instructions so that work is carried out in a controlled and consistent manner. Comply with all relevant safety, quality, and environmental management policies, procedures, and controls to ensure a healthy and safe work environment. Information Security Comply with all relevant information security practices and standards to ensure data integrity and confidentiality. JOB SPECIFICATIONS Academic and professional qualifications Bachelor’s degree in Computer Science, Software Engineering, Artificial Intelligence, Data Science, Computer Engineering, or a related field. Professional certifications in Machine Learning, Generative AI, MLOps, or Cloud Computing are considered an advantage. Years and Nature of Experience 2–4 years of relevant experience in applied AI/ML engineering, LLM application development, agentic systems, model evaluation, or AI solution deployment.