Gen AI Architect @ Edison, NJ (Hybrid 3 days onsite a week)

Saksoft

Edison, NJ

Contract

Senior

10 days ago

Job Description

About the Role

The position of Gen AI Architect is based in Edison, NJ, offering a hybrid work model with 3 days onsite per week. The role involves designing and governing advanced generative AI solutions, integrating various RAG strategies, and collaborating with product, engineering, security, and governance teams to ensure responsible deployment and long-term AI product strategy.

Key Responsibilities

Architect advanced Gen AI solutions incorporating various RAG strategies such as Naive Multihop, Graph-based RAG, etc.
Define patterns for integrating AI agents and agentic workflows using Azure AI services.
Design disconnected architectures with Azure Service Bus, Blob Storage, Cosmos DB, and AI Search.
Architect solutions for vector search, hybrid retrieval, and semantic ranking systems.
Plan and govern the deployment of large language models via Azure OpenAI MCP servers.
Guide prompt engineering best practices including prompt chaining, dynamic prompt tuning, and prompts.
Lead evaluation of grounding data sources, embeddings strategy, and data pipelines.
Collaborate with Security and Governance teams to implement ethical AI and responsible deployment practices.
Work with product and engineering leadership on long-term AI product strategy.

Requirements

10 years in enterprise architecture.
3 years in AIML.
Deep knowledge of Gen AI patterns, RAG, Naive Graph, Agent RAG, semantic pipelines, chain of thought.
Experience with AI agents and agentic workflows using CrewAI.
Experience integrating and managing models using Azure OpenAI.
Strong Azure knowledge including Cosmos DB, App Service, Azure AI Search, Blob Storage, and Document Intelligence.
Familiarity with LLMOps, vector databases, embedding tuning, and custom knowledge ingestion.
Proven track record in system-wide design and AI system governance.