Senior Azure AI and MLOps Engineer for Customer Service AI Platform

Summary Please start your proposal with the word AZURE so I know you have read the full post. We are building an Azure based AI customer service platform for a retail environment and are looking for an experienced Azure AI and MLOps engineer or architect to review the architecture and work with us on selected parts of the implementation. The platform will support customer service teams handling both English and Arabic enquiries across ecommerce and retail operations. The solution will include two main areas. Customer Service AI Assistant The assistant will use company policies, FAQs, product information and support knowledge to answer customer service questions with supporting sources. The current direction includes: • Azure AI Search • Azure OpenAI and RAG • Microsoft Foundry • Azure Blob Storage • GitHub and CI/CD • Azure Monitor and Application Insights • Secure configuration and secrets management Customer Escalation Model We also want to build and operationalize an Azure Machine Learning model that identifies customer service cases with a higher risk of escalation. The model will use information such as ticket category, previous contacts, response time, complaint type and other service data. A major part of the project is not only building the models, but managing them properly in production. We need to cover areas such as: • Development, testing and production environments • Model and prompt versioning • Automated evaluation before deployment • CI/CD for AI changes • Monitoring response quality, latency, failures and cost • Model endpoint monitoring • Retraining and model version management • Logging and tracing • Deployment approval and rollback • Quality monitoring after release One important requirement is multilingual performance. The platform needs to work well for both English and Arabic customer enquiries. We want to properly test retrieval and response quality in both languages and identify situations where the AI performs well in English but quality drops for Arabic users. We also want the selected resource to identify at least one difficult production issue they have repeatedly seen in Azure AI, RAG or MLOps projects and help us design the solution to handle it properly. For the first stage, we want you to review the proposed architecture, point out weaknesses, recommend improvements and discuss which parts of the implementation make sense for us to work on together. We are looking for someone with practical production experience, not only someone familiar with the services at a high level. The work will be split between our team and the selected resource. There is potential for longer term collaboration if there is a good fit. A detailed technical brief will be shared with shortlisted candidates before the interview. Relevant experience: Azure Machine Learning Azure OpenAI Azure AI Search Microsoft Foundry RAG MLOps and GenAIOps GitHub Actions Azure Monitor and Application Insights Python Azure security and production deployment