Hitachi
Agentic Systems Architect
Dallas, Texas, United States · Hybrid · Full-time · Posted Oct 23, 2026
- Experience
- 10+ years in software engineering, platform engineering, or systems architecture; 2+ years shipping production LLM or agent applications.
About the job
Designs and deploys governed, secure, cost-controlled agentic systems that connect LLM infrastructure with enterprise data, analytics, and operations.
What you'll do
- Architect multi-agent systems, MCP integrations, semantic layers, knowledge graphs, and vector retrieval for business use cases.
- Build guardrails, evaluations, observability, cost controls, and secure deployment workflows for production agents.
- Enable safe reasoning over complex data, including text-to-SQL and supplier or parts relationship queries.
What they ask for
- Hands-on experience with multi-agent frameworks and production MCP servers.
- Production experience shipping LLM or agent applications and evaluation frameworks as release gates.
- Experience with lakehouses and either production vector databases or knowledge graphs.
- Python, REST API design, Terraform, CI/CD, and AWS; experience with observability and cloud cost optimization.
Nice to have
- Databricks experience, automotive or supply-chain expertise, or GraphRAG and multi-hop reasoning.
- Model routing and cost optimization, human-in-the-loop workflows, or prompt versioning and systematic evaluation.
Tools
LLMs, MCP, Databricks, OpenSearch, Pinecone, Python, Terraform, AWS
AI focus
Designs and deploys production multi-agent systems, including LLM orchestration, tool calling, memory, guarded reasoning, evaluation, and cost monitoring.
Model-written summary · the original posting is authoritative