GenAI Architect

You will define enterprise AI architecture for agentic systems, RAG patterns, and autonomous agent orchestration.

This role is on-site.

Responsibilities

• Design enterprise AI architecture incorporating Agentic RAG, Graph RAG, Master-Agent orchestration, and MCP-based extensibility.
• Lead end-to-end architecture for intelligent copilots, autonomous agents, and workflow orchestration systems using open-source and commercial LLMs.
• Architect scalable ingestion pipelines for PPT/PDF content using Unstructured.io, LayoutParser, and Vision-LLMs for complex layout understanding.
• Implement AI safety, grounding, and evaluation systems using DeepTeam and NVIDIA NeMo Guardrails to enforce policy constraints and reduce hallucinations.
• Provide technical leadership by creating reference architectures for AgentOps, RAG frameworks, and tool-calling patterns across engineering teams.

Required Skills

• 10+ years of experience in software architecture or AI engineering.
• Deep expertise in Agentic RAG, Graph RAG, and multi-agent orchestration.
• Strong proficiency with LLMs, Vision-LLMs (e.g., Florence-2, LLaVA), and diffusion models.
• Experience with MLflow for lineage tracking and Databricks ML/Unity Catalog for governance.
• Knowledge of tool-calling patterns and secure agent-to-system interactions.
• Familiarity with inference frameworks like vLLM, TGI, or Hugging Face pipelines.
• Understanding of evaluation frameworks such as DeepEval or Ragas.
• Ability to communicate complex AI strategies to technical and executive stakeholders.

Preferred Skills

• Experience with Databricks Vector Search and semantic retrieval.
• Knowledge of NVIDIA NeMo Guardrails and policy enforcement.