Technical practice
How would you design an end-to-end AI solution for a client's business process?
What to coverThe role involves identifying automation opportunities and delivering secure, production-grade solutions integrated with client technology stacks. Preparation suggestion: explain how you would establish the business objective, choose system components, plan integration and measure whether the solution improves the process.
When would you use prompt engineering, tools, model selection or conventional software instead of an AI agent?
What to coverThe posting explicitly calls for judgment across AI tooling approaches, including when to rely on skills rather than AI. Preparation suggestion: compare reliability, complexity, maintainability and business value when choosing an approach.
How would you design and troubleshoot a retrieval-augmented generation system?
What to coverThe role names RAG, and understanding the pattern is required; practical implementation is preferred, not required. Preparation suggestion: cover document preparation, retrieval and answer generation, then explain how you would investigate poor retrieval or unsupported answers.
How would you configure and use an MCP server in an AI solution?
What to coverExperience configuring and using MCP servers is listed as a requirement. Preparation suggestion: describe how a client would discover and invoke available capabilities, and how you would limit access and handle failed tool calls.
How would you integrate an AI application with a client's APIs securely?
What to coverThe posting requires building and consuming APIs and understanding OAuth, API keys and token management; deeper OAuth and OIDC expertise is preferred. Preparation suggestion: discuss authentication versus authorisation, credential handling, error handling and the trade-offs between relevant access methods.
How would you choose and use an LLM framework for an end-to-end solution?
What to coverHands-on experience with at least one framework such as LangChain, LlamaIndex or Google ADK is required; Google ADK experience is separately preferred. Preparation suggestion: explain a framework you know through a concrete design decision, including what it simplifies and where direct application code may be clearer.
How would you keep Python or JavaScript AI application code production-grade?
What to coverThe posting emphasises software engineering, testing, code quality and algorithmic thinking rather than proof-of-concept delivery alone. Preparation suggestion: describe code structure, tests for deterministic and model-dependent behaviour, and how you would review correctness before release.
How would you use AI coding assistants while setting prompt and developer-experience standards?
What to coverThe role includes AI-assisted development and establishing patterns, standards and tooling for internal and client developers. Preparation suggestion: explain how you would verify generated code and make team conventions reusable without sacrificing security or maintainability.
How would you secure an AI solution deployed on AWS, GCP or Azure?
What to coverThe posting requires experience on at least one major cloud provider and application of security principles to code and cloud environments. Preparation suggestion: walk through access control, secret handling and deployment configuration for a provider you know, explaining how you would check for security gaps.
How would you containerise, deploy and monitor an AI application in production?
What to coverBasic DevOps awareness sufficient to containerise, deploy and monitor solutions is required; Kubernetes, GitOps and Infrastructure as Code are preferred rather than required. Preparation suggestion: outline a release and monitoring approach, then describe how you would investigate a failing deployment or degraded service.
Behavioural practice
Tell me about a time you clarified an ambiguous business process before proposing automation.
What to coverThe role involves asking questions, gathering requirements and identifying opportunities for AI-driven optimisation. Preparation suggestion: use a real situation/action/result example that shows what you learned, how it changed the proposed approach and what outcome followed.
Describe a time you managed competing scope, schedule and budget pressures.
What to coverThe posting assigns responsibility for tracking project requirements, progress, budgets and dependencies. Preparation suggestion: give a real situation/action/result example covering how you prioritised, communicated trade-offs and managed the resulting change.
Tell me about a time you raised a blocker with a client or stakeholder.
What to coverProactive progress reporting and raising blockers are explicit expectations. Preparation suggestion: use a real situation/action/result example explaining when you escalated, what options you presented and how the issue was resolved.
Describe a time you explained a technical solution to a non-technical audience.
What to coverThe role calls for presenting technical concepts to non-technical stakeholders and upskilling client teams. Preparation suggestion: give a real situation/action/result example showing how you adapted the explanation, checked understanding and supported the audience's next steps.
Tell me about a time you learned a new AI technology and decided whether to adopt it.
What to coverThe posting values curiosity about emerging AI technology and the ability to inspire colleagues and clients. Preparation suggestion: use a real situation/action/result example covering how you evaluated its usefulness, shared what you learned and reached a practical decision.