AI Development & ML Engineering Prompts

Models, pipelines, evaluation, and the engineering of learning systems. 16 prompts in this category.

• Model Evaluation Plan BuilderBefore training, so evaluation is not an afterthought.
• Dataset Documentation BuilderBefore sharing a dataset, to prevent misuse and enable reproducibility.
• ML Pipeline Design ReviewerBefore building a pipeline that will be hard to debug in production.
• Model Card DrafterBefore deploying a model that stakeholders will rely on.
• Feature Engineering AdvisorBefore modeling, when feature selection is the leverage point.
• Training Failure DiagnosticFor training runs where guessing wastes GPU hours.
• Data Drift Detection PlanBefore a model degrades silently in production.
• Inference Optimization AdvisorWhen inference cost or latency is blocking production.
• Dataset Bias AuditorBefore training a model that will make decisions about people.
• Experiment Tracking PlanBefore experiments multiply and become unrepeatable.
• LLM Fine-Tuning Decision FrameworkBefore committing to fine-tuning that may not be the right tool.
• Labeling Guidelines WriterBefore sending data to annotators, to reduce label noise.
• Model Serving Architecture AdvisorBefore building serving infrastructure that will be hard to change later.
• Active Learning Strategy PlannerWhen labeling is expensive and you want the most signal per dollar.
• MLOps Maturity AssessorFor ML teams that need to prioritize infrastructure investment.
• Responsible AI Review ChecklistBefore deploying a model where the societal impact is non-trivial.

Browse every category →