Description

AI Product Lead | 7+ years | Switzerland | Regulated sector | Azure OpenAI, Azure ML & Microsoft Fabric; own AI and ML products end-to-end from problem framing to production adoption, where governance, responsible-AI, and executive communication matter as much as technical delivery. About the role: We are looking for a highly senior Principal AI Platform Engineer & MLOps Architect (Azure) to help shape and build the bank’s core AI platform capabilities from the ground up. This role will partner closely with Data Science and data platform teams to transform an already mature and governed Azure ecosystem into a scalable, enterprise-grade Cognitive Data Platform for real-time and generative AI use cases. Location: Geneva, Switzerland Key Responsibilities: AI Platform architecture & feature management • Design and operationalize an enterprise Feature Store within the Azure ecosystem, enabling Data Scientists to discover, version, govern, and reuse features across batch and near-real-time use cases. • Define the target architecture for offline and online feature serving, with strong focus on consistency, scalability, and low-latency access. • Mitigate training-serving skew by implementing robust feature materialization and synchronization patterns across analytical and production environments. • Establish reusable platform standards for feature engineering, feature publishing, and production ML consumption. Vector search, RAG & LLMOps: • Architect and scale vector database capabilities for enterprise AI and Generative AI use cases using Azure-native services. • Design and implement data chunking, embedding, metadata tagging, and retrieval pipelines to support high-quality Retrieval-Augmented Generation (RAG) solutions. • Evaluate and implement fit-for-purpose patterns across Azure AI Search, Azure Cosmos DB vector capabilities, and related services for semantic and hybrid search. • Contribute to the operationalization of LLM-backed services with focus on reliability, performance, and governance. Near-real-time pipelines & inference: • Build near-real-time ingestion pipelines using Azure-native streaming services such as Event Hubs, Stream Analytics, and/or Databricks Structured Streaming. • Design and implement production-grade inference pipelines capable of serving live model predictions with low latency and high throughput. • Deploy and manage online inference services through Azure Machine Learning endpoints and/or containerized platforms such as AKS. • Ensure production readiness through monitoring, alerting, resiliency, and scalable deployment patterns. Platform ownership & technical leadership: • Act as the lead technical bridge between Data Science, enterprise data governance, and platform engineering teams. • Translate advanced AI experimentation into modular, secure, and production-ready MLOps solutions. • Provide architectural direction, engineering standards, and hands-on guidance for AI platform buildout. • Mentor technical stakeholders on platform best practices, operational excellence, and sustainable delivery models. Security, governance & ways of working: • Ensure all AI platform components align with enterprise security controls, data classification policies, and governance requirements. • Apply best practices across secrets management, access control, encryption, auditability, and compliant data usage. • Promote Infrastructure as Code, CI/CD automation, and repeatable deployment standards across the AI platform stack. • Work in close collaboration with cross-functional stakeholders in an agile, delivery-focused environment. First 90 days / expected impact • Month 1: Assess the existing Azure data platform, governance model, and Data Science workflows; produce the target architecture blueprint for Feature Store and vector platform capabilities. • Month 2: Launch the core platform foundations, including managed Feature Store components and vector search infrastructure aligned to production use cases. • Month 3: Deliver the first end-to-end near-real-time ingestion and inference pipelines, enabling production-grade AI use cases with measurable latency and reliability targets. Experience • 7+ years of experience in Data Engineering, Cloud Architecture, Platform Engineering, or MLOps. • At least 3 years of recent experience building and productionizing machine learning platforms, inference systems, or LLM infrastructure. • Proven track record designing platform components such as Feature Stores, vector search backends, or enterprise AI/ML infrastructure from scratch. • Strong experience delivering production solutions in Microsoft Azure environments. Skills: • Deep hands-on knowledge of Azure Machine Learning, including workspaces, managed services, online endpoints, and MLOps patterns. • Strong experience with Azure AI Search, Azure Cosmos DB vector capabilities, and/or similar vector database technologies. • Experience with streaming and real-time data technologies such as Azure Event Hubs, Azure Stream Analytics, Azure Functions, and Azure Databricks. • Solid understanding of Azure Data Lake Storage Gen2, Microsoft Fabric / OneLake, and enterprise data platform integration patterns. • Strong coding skills in Python and PySpark. • Proven experience with Infrastructure as Code using Terraform and/or Bicep. • Good command of CI/CD practices using Azure DevOps and/or GitHub Actions. • Strong architectural thinking, with ability to combine strategy, hands-on engineering, and delivery ownership. • Clear communication skills and confidence working with both technical and non-technical stakeholders. • Fluent in English, spoken and written. Nice to have: • Experience in private banking, financial services, or other highly regulated environments. • Exposure to enterprise data governance, model risk controls, and secure AI deployment practices. • Familiarity with multi-tenant ML platform design and low-latency online serving architectures. • Experience mentoring Data Scientists or engineering teams on production MLOps best practices. Your Data: By submitting your resume, you agree to the retention and use of your personal data by TSG for recruitment purposes, including sharing with our clients in the context of your application.