AI/ML Solution Architect

Key Responsibilities

1. AI/ML Solution Architecture

• Define end-to-end architectures for enterprise AI/ML solutions covering data ingestion, processing, feature engineering, model development, inference, application integration, monitoring, and governance.
• Translate business and operational requirements into scalable technical architectures.
• Select appropriate AI/ML frameworks, cloud services, infrastructure, and deployment patterns.
• Establish architecture standards, reference architectures, design patterns, and technology roadmaps.
• Evaluate emerging AI/ML technologies and determine their suitability for business requirements.

2. Computer Vision and Edge AI

• Design end-to-end camera-to-edge-to-cloud-to-application workflows.
• Architect real-time image and video analytics solutions.
• Work with YOLO, OpenCV, CNN/Transformer-based models, image classification, object detection and tracking, segmentation, OCR, and anomaly detection.
• Optimize models for cloud and edge deployment.
• Design GPU-enabled inference environments using containers and Kubernetes where appropriate.
• Integrate computer vision solutions with industrial systems, IoT platforms, APIs, dashboards, and enterprise applications.
• Address video bandwidth, latency, inference performance, scalability, and edge connectivity challenges.

3. Azure Cloud Architecture

• Design secure, scalable, highly available, and cost-effective Azure cloud architectures.
• Develop hybrid cloud and edge architectures for AI inference close to industrial assets and data sources.
• Define cloud infrastructure and deployment patterns that meet performance and reliability requirements.
• Support cloud-native AI/ML application development and enterprise integration.

4. MLOps and AI Lifecycle Management

• Establish enterprise-grade MLOps architecture and practices.
• Define model versioning, experiment tracking, model registry, and artifact management.
• Design monitoring for model performance, data drift, concept drift, data quality, infrastructure health, latency, and throughput.
• Enable automated retraining and model lifecycle management.
• Integrate ML pipelines with Azure DevOps, GitHub, CI/CD, and Infrastructure-as-Code practices.

5. Data Architecture

• Define data architectures supporting AI/ML workloads from ingestion to consumption.
• Design batch and real-time data pipelines.
• Establish ingestion patterns using APIs, event streams, IoT telemetry, databases, and files.
• Architect solutions for structured, semi-structured, and unstructured data.
• Ensure data architectures meet scalability, security, availability, and AI/ML requirements.

6. API and Integration Architecture

• Design RESTful APIs and event-driven integration architectures.
• Develop architectural patterns for microservices-based AI/ML applications.
• Use Azure API Management and other integration services where appropriate.
• Integrate AI/ML services with enterprise applications, IoT/OT platforms, mobile and web applications, and Power Platform.
• Define secure and reliable application integration patterns.

7. Application and Platform Architecture

• Design cloud-native applications using microservices and containerized workloads.
• Define scalable backend services supporting AI/ML applications.
• Establish synchronous and asynchronous processing patterns.
• Define caching, messaging, database, and service-discovery strategies.
• Ensure applications support horizontal scaling and high availability.
• Guide development teams on implementation standards and architectural best practices.

8. Security and Governance

• Incorporate security by design across AI, data, API, and cloud architectures.
• Define data protection, encryption, and access control mechanisms.
• Address AI/ML governance, responsible AI, model security, and auditability.
• Ensure compliance with enterprise security and regulatory requirements.

9. Productionization and Enterprise Scale

• Assess PoCs and identify the architectural changes required for production readiness.
• Identify scalability, reliability, security, and operational gaps.
• Establish production deployment and operational support patterns.
• Design solutions that support large numbers of users, devices, cameras, industrial assets, or sites.
• Define Service Level Agreements (SLAs), Service Level Objectives (SLOs), and non-functional requirements.
• Optimize compute, storage, networking, and AI inference cos…