Technical practice
Architecting Agentic Workflows
What to coverExplain how you design multi-agent systems using LangGraph to handle complex planning and tool-calling tasks. Discuss how you manage state, memory, and error recovery in these workflows to ensure reliability in production.
Optimizing RAG Pipelines
What to coverDescribe your approach to improving RAG performance beyond basic retrieval. Address specific techniques for chunking strategies, embedding selection, and the implementation of re-ranking mechanisms to increase precision.
Production-Grade Python and API Design
What to coverDiscuss your process for building scalable, production-ready AI services using FastAPI. Focus on handling asynchronous operations, managing concurrency, and ensuring robust error handling for model inference endpoints.
Vector Database Selection and Scaling
What to coverCompare different vector databases like Pinecone, Milvus, or Qdrant based on performance, scalability, and cost. Explain how you choose the right indexing strategy for a specific use case to balance search latency and accuracy.
LLM Evaluation Frameworks
What to coverHow do you systematically evaluate LLM outputs in a production environment? Suggest using frameworks like Ragas or LangSmith to measure metrics such as faithfulness, answer relevance, and context precision.
Fine-Tuning Strategies (LoRA/QLoRA)
What to coverExplain the trade-offs between full fine-tuning and parameter-efficient methods like LoRA or QLoRA. Discuss when it is appropriate to fine-tune a model versus relying on RAG or prompt engineering.
MLOps and Model Deployment
What to coverDescribe your workflow for deploying models using Docker and cloud-native tools like SageMaker or MLflow. Focus on versioning, monitoring for model drift, and automating the CI/CD pipeline for AI models.
Classical ML vs. GenAI Integration
What to coverDiscuss scenarios where traditional machine learning models (scikit-learn) are more effective than LLMs. Explain how you integrate these different paradigms into a single, cohesive product architecture.
Behavioural practice
Leading and Mentoring Technical Teams
What to coverDescribe a situation where you had to mentor a junior engineer or lead a technical project. Focus on how you balanced hands-on development with code reviews and design guidance to ensure team growth and project success.
Translating Business Requirements to AI Solutions
What to coverProvide an example of how you translated a vague business problem into a concrete AI product. Explain how you communicated technical trade-offs to non-technical stakeholders to manage expectations.
Handling Production Failures
What to coverDescribe a time when a production AI system failed or performed unexpectedly. Detail the steps you took to diagnose the issue, the immediate mitigation, and the long-term architectural changes you implemented to prevent recurrence.
Managing Technical Debt in AI Projects
What to coverExplain how you balance the need for rapid prototyping in AI with the requirement for maintainable, production-grade code. Give an example of how you prioritized refactoring or infrastructure improvements alongside new feature development.