AI Engineer

NTT DATA•Parsipanny, NJ

14d•Remote

About The Position

We are currently seeking a professional to join our team in New Jersey, United States. This role focuses on being a Master Coordinator, domain agents, semantic retrieval, knowledge graphs, data foundations, and agent quality. The AI Engineer will be responsible for advanced Python, agent services, multi-step workflows, orchestration state, tool invocation, fallback handling, and reusable agent patterns. They will also work with standard agent and tool contracts, agent registration, deterministic and LLM-assisted routing, multi-domain execution, and model and platform adapters. Data engineering responsibilities include advanced SQL and BigQuery, source discovery, data-gap analysis, data dictionaries, business-key validation, reconciliation, and assessing quality, freshness, lineage, and source-of-truth. AI and Retrieval expertise is crucial, including deep RAG expertise, retrieval and reranking, grounding, prompt/context assembly, citation support, hallucination reduction, and cross-domain synthesis. Knowledge Graph skills involve RDF, SPARQL, ontology modeling, SHACL, Stardog, virtual graphs, relational-to-semantic mappings, semantic versioning, and graph promotion. Cloud and Infrastructure experience with GCP, GKE, BigQuery, GCS, agent-runtime deployment, semantic-platform connectivity, environment configuration, and secure service identities is required. CI/CD and Release responsibilities include agent and knowledge-graph CI/CD, automated evaluation gates, and versioned prompts, tools, mappings, ontologies, queries, and deployment definitions. Testing and Quality involve agent evaluation for accuracy, relevance, groundedness, completeness, hallucination, latency, and cost, as well as benchmark scenarios and gold-answer datasets. Observability and Operations require end-to-end tracing across coordinator, agent, tool, semantic, and data layers, LLM observability (e.g., Langfuse), Grafana reporting, and model comparisons. Security and Governance include identity-aware retrieval, least-privilege data access, cross-domain guardrails, trace and prompt-data protection, and source attribution with complete auditability. Tech Lead Expectations include owning orchestration, semantic, and data-foundation design; remaining hands-on; creating reusable patterns; mentoring engineers; and driving evaluation and production readiness.

Requirements

Advanced Python
Agent services
Multi-step workflows
Orchestration state
Tool invocation
Fallback handling
Reusable agent patterns
Standard agent and tool contracts
Agent registration
Deterministic and LLM-assisted routing
Multi-domain execution
Model and platform adapters
Advanced SQL and BigQuery
Source discovery
Data-gap analysis
Data dictionaries
Business-key validation
Reconciliation
Quality, freshness, lineage, and source-of-truth assessment
Deep RAG expertise
Retrieval and reranking
Grounding
Prompt/context assembly
Citation support
Hallucination reduction
Cross-domain synthesis
RDF, SPARQL, ontology modeling
SHACL, Stardog
Virtual graphs
Relational-to-semantic mappings
Semantic versioning
Graph promotion
GCP, GKE, BigQuery, GCS
Agent-runtime deployment
Semantic-platform connectivity
Environment configuration
Secure service identities
Agent and knowledge-graph CI/CD
Automated evaluation gates
Versioned prompts, tools, mappings, ontologies, queries, and deployment definitions
Agent evaluation for accuracy, relevance, groundedness, completeness, hallucination, latency, and cost
Benchmark scenarios and gold-answer datasets
End-to-end tracing across coordinator, agent, tool, semantic, and data layers
Langfuse or equivalent LLM observability
Grafana reporting and model comparisons
Identity-aware retrieval
Least-privilege data access
Cross-domain guardrails
Trace and prompt-data protection
Source attribution and complete auditability