Assistant Vice President- GenAI Enterprise Architect

Role Overview :

We are seeking a highly experienced and strategic AVP - GenAI Architect to lead the architecture, design, and deployment of enterprise-grade Generative AI solutions using Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), AI agents, and emerging AI technologies.

The role will be responsible for defining and executing the organization's GenAI technology strategy, translating business opportunities into scalable AI solutions, and establishing architecture standards and engineering practices for production-grade GenAI implementations.

Key Responsibilities :

• Define and lead the architecture of enterprise-grade Generative AI solutions.

• Design scalable solutions leveraging LLMs, RAG, AI/LLM agents, multi-agent architectures, conversational AI, AI copilots, enterprise search, document intelligence, summarization, classification, knowledge assistants, and workflow automation.

• Develop reference architectures, design patterns, standards, and reusable components for GenAI applications.

• Evaluate build-versus-buy decisions for AI platforms and capabilities.

• Design solutions that support scalability, reliability, performance, observability, and cost optimization.

• Ensure architecture decisions align with enterprise technology strategy and long-term platform objectives.

• Establish architectural principles for production-grade GenAI development and deployment.

• Evaluate and select appropriate foundation models based on use case requirements, including accuracy, context-window, latency, cost, security, data privacy, availability, and fine-tuning capabilities.

• Design model-routing and model-selection strategies.

• Evaluate hosted APIs, managed foundation models, and self-hosted/open-source models.

• Establish standards for model evaluation, benchmarking, versioning, and lifecycle management.

• Assess opportunities for fine-tuning, instruction tuning, prompt optimization, and retrieval-based approaches.

• Design and implement enterprise-grade RAG architectures.