Model drift is the gradual decline in a model’s predictive quality during operation because reality changes compared with the training data. A distinction is made between changes in the input data (data drift) and changes in the relationship between inputs and outcomes (concept drift). Example: a fraud detection model misses new fraud schemes. Remedies include continuous monitoring of metrics and regular retraining.
Model robustness describes how reliably an AI model performs when inputs are noisy, incomplete, unusual or deliberately manipulated. A robust model still delivers stable results, such as image recognition in poor lighting or a language model handling typos. Robustness is assessed with stress tests, diverse training data and tests against deliberate deception attempts, and it is particularly important in safety-critical applications.
Training data are the examples from which an AI model learns patterns: large text collections for language models, labelled images for image models. Their quality, diversity and recency largely determine how reliable and fair a model is and how up to date its knowledge will be. For organisations it also matters whether a provider uses the customer data they enter for further training.
MLflow is an open-source platform for managing the entire life cycle of machine learning models, originally introduced by Databricks in 2018. It records experiments with their parameters and metrics, manages model versions in a registry and simplifies deployment. More recent versions also support tracing and evaluating applications built on language models. Teams can therefore trace which model was produced with which data and settings.
Accuracy describes how often an AI model produces correct results; stability describes how reliably that performance holds up – with slightly altered inputs, new data or over time. A model can be very accurate in testing yet unstable if merely rephrasing a question leads to a different answer. For real-world use, both properties are checked with fixed test cases, cross-validation and ongoing monitoring.
