Product Architect - AI/ML

Product Architect - AI/ML
Location: India/Offshore
Experience: 12-15 years (Minimum 6-7 years in AI/ML Product Architecture)

Role Overview
Define and drive the technical product vision for enterprise AI/ML platforms in healthcare, translating business requirements into scalable architectures while ensuring delivery excellence and long-term product sustainability. This role requires hands-on leadership, flexible collaboration across time zones, and the ability to mentor engineering teams.
Product & Technical Strategy
• Architect end-to-end product solutions for processing clinical records, claims, and healthcare documentation
• Design hybrid Azure/on-premises architectures supporting multi-tenant SaaS and enterprise deployment models
• Establish product architecture principles, design patterns, and technology stack decisions
• Evaluate build-vs-buy decisions for AI capabilities, Drive technical feasibility assessments and rapid prototyping for new product features
• Own non-functional requirements: performance, security, compliance (HIPAA), scalability, reliability
• Present architectural proposals and technical roadmaps to leadership through clear presentations and documentation AI/ML Product Engineering
• Design and implement fine-tuning pipelines for domain-specific LLMs (e.g. Llama, Mistral, Phi-3) on medical datasets
• Architect MLOps frameworks enabling continuous model improvement: training, evaluation, deployment, monitoring
• Build “built-in-intelligence-products” using advanced AI models.
• Implement distributed training infrastructure on Azure ML with GPU optimization (RTX5090/4090/A100/H100 clusters)
• Create model evaluation frameworks with domain-specific metrics and quality benchmarks
• Develop synthetic data generation capabilities for model training and testing
• Establish model versioning, A/B testing, and rollback strategies for production deployments Platform & Infrastructure
• Design microservices architectures with Azure Kubernetes Service, Service Bus, Event Hubs for event-driven workflows
• Build scalable data pipelines using Azure Data Factory, Synapse Analytics, Databricks for high-volume processing
• Implement CI/CD automation through Azure DevOps with comprehensive testing and deployment gates
• Architect observability solutions: monitoring, logging, alerting, performance analytics
• Design hybrid cloud integration patterns connecting Azure services with on-premises systems
• Manage data architecture: Azure SQL, Cosmos DB, blob storage, data lakes with lifecycle policies Delivery Excellence & Hands-On Leadership
• Lead proof-of-concept execution from inception to production-ready solutions within defined timelines
• Conduct rigorous code reviews ensuring adherence to standards, performance optimization, and maintainability
• Roll up sleeves for critical implementation work: debugging production issues, optimizing model performance, refactoring complex modules
• Deliver technical demos to internal stakeholders and customers showcasing product capabilities
• Establish engineering excellence standards: code quality, testing frameworks, peer reviews, documentation
• Create comprehensive technical documentation: architecture decision records, API contracts, deployment guides
• Identify and systematically address technical debt across AI models, infrastructure, and codebase Team Development & Collaboration
• Mentor and develop junior engineers and fresh graduates through pair programming, and technical guidance
• Review and provide constructive feedback on designs, code, and technical approaches from team members
• Foster engineering culture emphasizing quality, innovation, and continuous learning
• Collaborate with US-based teams requiring flexibility for meetings during early morning or late evening IST hours
• Bridge communication between offshore development teams and US-based product/business stakeholders • Translate business requirements from US teams into actionable technical tasks for India-based engineers
• Participate in cross-timezone planning, sprint reviews, and architecture discussions