Agent Developer - Knowledge Graph Specialize

The specialist agent developer focused on knowledge graph design, ontology modelling, graph-based retrieval, and grounding AI agent outputs with structured enterprise knowledge.

Responsibilities:

• Design and build ontology-driven knowledge graphs for domains, requirements, code assets, design artifacts, and enterprise knowledge sources.
• Model entities, relationships, taxonomies, hierarchies, inheritance rules, and metadata needed for agent reasoning and retrieval.
• Implement graph-based retrieval, hybrid RAG, semantic search, and context-ranking patterns to improve grounding and explainability.
• Integrate graph databases with agent workflows, prompt templates, and evaluation datasets.
• Validate knowledge quality, graph consistency, relationship accuracy, lineage, and provenance of retrieved context.
• Collaborate with domain, backend, and agent teams to operationalize reusable knowledge models and graph-backed guardrails.

Requirements:

• Bachelor's or master's degree in computer science, AI, data engineering, or a related discipline.
• 6+ years of total experience, including 2+ years in agent development, knowledge graphs, ontology modelling, or semantic retrieval.
• Strong Python programming expertise.
• Hands-on experience with graph databases, graph query languages, ontology design, embeddings, vector search, and RAG patterns.
• Experience integrating knowledge graphs with LLM applications, agent workflows, and enterprise knowledge sources.
• Familiarity with evaluation frameworks, grounding-quality checks, data lineage, provenance, and AI guardrails.

Mandatory Tech Stack:

• Language: Python.
• Agent Frameworks: LangGraph / LangChain or equivalent agent-orchestration frameworks.
• LLM APIs: OpenAI / Anthropic / Bedrock / Azure OpenAI or equivalent foundation-model APIs.
• Graph DB: Amazon Neptune / Neo4j and query languages such as Cypher, Gremlin, or SPARQL.
• Knowledge Modeling: Ontology design, taxonomies, hierarchical inheritance, entity/relationship modelling, metadata, lineage, and provenance.
• Retrieval: Vector databases such as Pinecone/FAISS, embeddings, semantic search, hybrid RAG, and grounding-quality checks.