Sr. AI Architect

We are looking for a Senior AI Architect to design and scale enterprise AI/ML platforms that bridge business needs with production-ready solutions. The role focuses on AI architecture, MLOps, model deployment, observability, and platform scalability rather than core model development.

Responsibilities:

• Design scalable AI/ML platforms, including data pipelines, model deployment, serving, and monitoring.
• Define architectures for real-time inference, recommendation systems, NLP/CV, and batch processing.
• Build cloud-native, Kubernetes-based, microservices-driven AI systems.
• Architect production-grade ML serving with A/B testing, canary releases, rollback strategies, and optimization techniques.
• Establish end-to-end MLOps practices: CI/CD, model versioning, feature stores, experiment tracking, and reproducibility.
• Implement ML observability, monitoring, alerting, and performance dashboards.
• Provide technical leadership through architecture reviews, mentoring, and AI governance.
• Evaluate emerging AI technologies including LLMs, RAG, multimodal AI, and distributed training infrastructure.

Requirements:

• 12+ years in software engineering with 6+ years in AI/ML architecture and production deployment.
• Strong experience building enterprise-scale ML platforms and distributed systems.
• Expertise in Python, PyTorch/TensorFlow, Kubernetes, Docker, MLflow/Kubeflow, Kafka, Redis, and cloud platforms (AWS/GCP/Azure).
• Deep understanding of MLOps, model serving, feature stores, monitoring, and production ML patterns.
• Experience with LLM architectures, vector databases, model optimization, and GPU-based deployments.
• Strong communication, leadership, and architectural decision-making skills.

Nice to Have:

• Experience in regulated industries (finance/healthcare).
• Knowledge of Edge AI, federated learning, NLP, CV, recommendation systems, or time-series ML.
• Master's or PhD in Computer Science, AI/ML, or a related field.

Skills

Docker, Kubeflow, Kubernetes, MLOps, MLflow, Python, feature store, microservices, model serving