Azure Classification Model Build
Budget: ₹600 – ₹1,500 INR
The project centres on creating production-ready classification models within the Azure ecosystem. I already have an Azure subscription and the raw, labelled data is stored in Blob Storage; what I now need is an end-to-end machine learning workflow that trains, evaluates, and then serves a robust classifier through Azure Machine Learning.
Key expectations
• Model focus: machine learning models, specifically classification.
• Tooling: Azure Machine Learning workspace, Python (preferred), and any relevant libraries such as scikit-learn, TensorFlow, or PyTorch.
• Lifecycle: data preprocessing, feature engineering, hyper-parameter tuning, cross-validation, and automated ML logging for reproducibility.
• Deployment: register the best model, deploy it to an Azure Container Instance or AKS endpoint, and expose a simple REST API.
• Handoff: clear documentation, Jupyter notebooks or scripts, and a brief walkthrough so I can retrain or update the model later.
If you have delivered classification solutions on Azure before and can demonstrate clean, well-commented code plus solid performance metrics (accuracy, precision-recall, confusion matrix), I’d be happy to discuss the dataset and timelines in more detail.
Key expectations
• Model focus: machine learning models, specifically classification.
• Tooling: Azure Machine Learning workspace, Python (preferred), and any relevant libraries such as scikit-learn, TensorFlow, or PyTorch.
• Lifecycle: data preprocessing, feature engineering, hyper-parameter tuning, cross-validation, and automated ML logging for reproducibility.
• Deployment: register the best model, deploy it to an Azure Container Instance or AKS endpoint, and expose a simple REST API.
• Handoff: clear documentation, Jupyter notebooks or scripts, and a brief walkthrough so I can retrain or update the model later.
If you have delivered classification solutions on Azure before and can demonstrate clean, well-commented code plus solid performance metrics (accuracy, precision-recall, confusion matrix), I’d be happy to discuss the dataset and timelines in more detail.