What are the responsibilities and job description for the Agentic AI / Semantic Solutions Architect position at Prudent Technologies and Consulting?
Job Title : Agentic AI / Semantic Solutions Architect
Location : Atlanta, Georgia, USA
Duration: Long Term
Experience : 15 Years
Role Overview
We are seeking a highly skilled Agentic AI / Semantic Solutions Architect to design and prototype advanced agent-layer architectures that operate on enterprise semantic data platforms. This role sits at the intersection of LLM orchestration, knowledge graphs, and semantic data modeling, focusing on building POC-level intelligent agent solutions rather than production-scale systems.
The ideal candidate will have deep expertise in agent-based AI systems, GraphRAG architectures, and context engineering, with the ability to design frameworks where autonomous agents can effectively interpret and reason over structured knowledge.
Key Responsibilities
- Architect and design agentic AI workflows that consume outputs from semantic layers, including knowledge graphs, ontologies, and metadata catalogs
- Develop and prototype GraphRAG pipelines that combine graph traversal with vector-based retrieval for accurate, domain-grounded responses
- Define and implement context engineering strategies, including metadata injection, chunking, and semantic optimization for LLM prompts
- Design and build Model Context Protocol (MCP) server patterns to enable seamless interaction between agents and semantic data systems
- Develop LLM orchestration workflows using frameworks such as LangChain, LangGraph, LlamaIndex, or AutoGen
- Build pipelines for automated metadata extraction and semantic tagging using NLP and LLM-based approaches
- Collaborate with Semantic Data Architects to ensure ontologies and graph structures are optimized for agent traversal and querying
- Prototype agent-based solutions for business use cases such as:
- Credit risk analysis
- Customer data onboarding workflows
Mandatory Skills
- Strong expertise in Agentic AI architecture (multi-agent systems, tool usage, planning loops)
- Hands-on experience with GraphRAG design (hybrid graph vector retrieval systems)
- Experience in LLM orchestration frameworks:
- LangChain, LangGraph, LlamaIndex, or AutoGen
- Deep understanding of context engineering techniques (chunking, windowing, semantic compression)
- Experience designing and integrating Model Context Protocol (MCP)
- Strong knowledge of semantic systems such as:
- Knowledge graphs
- Ontologies
- Metadata-driven architectures
Nice to Have Skills
- Experience with Google Vertex AI (Agent Builder / Search)
- Knowledge of Google Cloud Platform Spanner Graph
- Familiarity with metadata platforms like Collibra or Google Dataplex
- Experience with vector databases:
- Pinecone, Weaviate, pgvector, Vertex AI Vector Search
- Prior experience in regulated domains such as financial services or legal systems