What are the best practices for model versioning?
We are currently using folder names like 'model_v2_final_final' and it is becoming a disaster. How do professional teams handle model versioning? Should the versioning be tied to the git hash of the c...
How to handle sensitive user data in automated pipelines?
We have strict requirements for handling user data, but we also need to retrain our models on this data. How do we keep our pipelines compliant and secure without blocking our agile deployment goals? ...
What is the best way to handle model explainability?
Our users are asking why a model made a specific prediction. We don't have a good answer. How do you integrate explainability tools like SHAP or LIME into a production pipeline? Is it possible to gene...
What tools are recommended for tracking experiments in an agile setup?
Our data science team is having a hard time keeping track of model versions and experimental results. We want to move towards a more agile process where experiments are reproducible and easy to share....
How do you manage resource allocation for training vs inference?
Our cloud bills are getting out of control because we are running expensive GPU instances for both training and inference. In an agile MLOps environment, how do you dynamically balance these resources...
What are the common compliance requirements for ML?
We are entering a regulated industry and I am worried about the compliance requirements for our ML models. What do we need to track? Auditing? Bias reporting? Fairness audits? I want to make sure our ...
How to structure an agile team for ML projects?
Our current structure has data scientists in one silo and software engineers in another. It's making it impossible to do true agile MLOps. How do successful companies organize their teams? Should ever...
How to effectively manage feature stores in a scalable way?
Feature stores are a hot topic, but when is it actually worth the cost and complexity to implement one? We are growing our number of models, and the code for feature engineering is getting repetitive ...
What role does DVC (Data Version Control) play in MLOps?
I keep hearing about DVC in the context of agile MLOps and scalable deployments. Is it actually necessary for a team that is already using Git? I understand it helps with large files, but does it make...
When should I use a CNN versus an MLP?
I am working on a classification task for tabular data, and I am wondering if I should try a convolutional neural network or stick to a simple multi-layer perceptron. I usually see CNNs applied to ima...
