Senior AI Architect – Generative & Agentic Systems Lead

Virtusa Corporation; Tampa (FL); On-site; USD 120,000 - 190,000; Full time

Job summary

Virtusa Corporation seeks a Senior AI Developer to lead Generative AI, Agentic AI architectures and enterprise RAG pipelines. You will design, build and deploy scalable LLM applications and autonomous multi-agent systems across complex enterprise challenges.

Responsibilities include architecting foundation-model apps with commercial and open-source models, implementing prompt patterns, context management, PEFT, quantization, and robust evaluation guardrails, and guiding tool use and

Qualifications

• Experience designing and deploying large-scale LLM-powered solutions for enterprise use.
• Hands-on with multi-agent workflows and retrieval-augmented generation pipelines.
• Strong background in prompt engineering, structured outputs, and validation schemas.

Responsibilities

• Generative AI & LLM engineering and deployment.
• Architect, fine-tune and deploy foundation model apps using commercial and open-source models.
• Design robust prompt patterns, context/window management, and validation schemas.
• Implement optimization techniques (PEFT/LoRA, quantization) for latency and cost.
• Design and orchestrate deterministic autonomous multi-agent workflows.
• Enable tool use, planning, memory management, and HITL gates; adopt MCP & OpenAPI-driven tooling.

Skills

Generative AI

LLM Engineering

Agentic AI

Prompt engineering

JSON/Pydantic validation

Evaluation guardrails

Tools

OpenAI API

Anthropic API

Vertex AI

LangGraph

CrewAI

AutoGen

Semantic Kernel

Llama

Mistral

Qwen

Job description

Virtusa Corporation seeks a Senior AI Developer to lead Generative AI, Agentic AI architectures and enterprise RAG pipelines. You will design, build and deploy scalable LLM applications and autonomous multi-agent systems across complex enterprise challenges.

Responsibilities include architecting foundation-model apps with commercial and open-source models, implementing prompt patterns, context management, PEFT, quantization, and robust evaluation guardrails, and guiding tool use and