Requirements
▶ Education
– Bachelor’s degree in Computer Science, Software Engineering, Data Engineering, Information Technology, or a related discipline.
– Postgraduate qualification in Artificial Intelligence, Machine Learning, or Data Science is an advantage.
▶ Professional Certifications
– Certification in a major AI or cloud platform (Microsoft Azure AI Engineer, AWS Machine Learning, Google Cloud Professional ML Engineer, or equivalent).
– Developer-level certifications in Python, cloud application development, or integration platforms are advantageous.
– Recognised training in Agentic AI frameworks, RAG architecture, or LLM application security is an asset.
▶ Experience
– 4–7 years of experience in software development, automation, data engineering, AI/ML, or enterprise application integration.
– Hands-on experience in building, configuring, testing, and deploying AI agents or GenAI-based applications.
– Experience with AI/GenAI platforms such as OpenAI API, Microsoft Copilot Studio, Amazon Bedrock, Google Vertex AI, or similar.
– Familiarity with Agentic AI frameworks such as LangChain, LangGraph, CrewAI, AutoGen, Semantic Kernel, or similar.
– Demonstrated experience integrating applications with enterprise systems, databases, document repositories, and workflow tools.
– Exposure to production support, monitoring, and incident resolution for deployed solutions.
▶ Key Skills & Attributes
– Strong practical coding ability, with clean, testable, and well-documented implementation.
– Structured debugging and root-cause analysis for probabilistic systems where failures are not always reproducible.
– Clear judgement on when a prototype is genuinely production-ready and when it is not.
– Ability to work directly with business users to refine requirements and validate outputs.
– Disciplined approach to security, data protection, and access control in every build.
– Curiosity and self-directed learning in a technology area that changes month to month
Description
The AI Engineer / Agent Developer is the hands-on builder of the Group’s Agentic AI capability, turning approved use cases into working, reliable agents that operate safely against real enterprise systems and data.
The role spans rapid prototyping and production engineering — designing prompts and retrieval strategies, integrating agents with core platforms and tools, and instrumenting them so that accuracy, reliability, cost, and exception handling can be measured and improved. Success is judged not by demonstrations but by agents that hold up in daily business use under enterprise standards of security and control.
▶ Agent Development & Engineering
– Build, configure, and test AI agents against defined business use cases, translating solution blueprints into working implementations.
– Work hands-on across prompt engineering, retrieval-augmented generation (RAG), workflow automation, API development, and tool integration.
– Implement agent reasoning patterns, tool selection logic, memory and context management, and structured output handling.
– Design and implement guardrails — input validation, output constraints, confidence thresholds, and safe failure behaviour.
– Apply disciplined engineering practice: version control, code review, environment separation, automated testing, and CI/CD pipelines.
▶ Prototyping & Production Deployment
– Develop rapid prototypes that prove or disprove feasibility quickly, with clear articulation of assumptions and limitations.
– Support the transition of validated prototypes into production deployment, including hardening, performance tuning, and operational documentation.
– Prepare release artefacts, runbooks, and support handover materials for infrastructure and application support teams.
– Contribute to shared libraries, reusable components, prompt templates, and evaluation harnesses that accelerate future builds.
▶ Enterprise Systems Integration
– Connect AI agents with enterprise systems, documents, databases, and workflow tools, including ERP, CRM, HRMS, procurement, and document repositories.
– Build and consume secure APIs and integration services, managing authentication, rate limits, error handling, and retry logic.
– Implement human-in-the-loop approval steps and escalation routes so that agent actions remain reviewable and reversible.
– Coordinate with enterprise application owners on data contracts, sandbox access, regression testing, and release windows.
▶ Performance Monitoring & Quality Assurance
– Monitor agent performance, accuracy, reliability, latency, and exception handling in both test and production environments.
– Build evaluation datasets and automated test suites to detect regression when prompts, models, or upstream data change.
– Investigate failures and unexpected behaviour to root cause, and implement corrective changes with documented evidence of improvement.
– Track token consumption and inference cost, and optimise model selection, context size, and caching accordingly.
▶ Collaboration, Security & Documentation
– Work closely with the AI / Agentic AI Lead, data engineers, application specialists, and business users throughout the delivery cycle.
– Apply cybersecurity and data protection requirements in every build, including data classification, secrets management, and least-privilege access.
– Maintain clear technical documentation covering architecture, prompts, integrations, known limitations, and support procedures.
– Support user enablement by demonstrating capability, gathering structured feedback, and iterating on real-world usage.