Senior Knowledge Graph & Reasoning Engineer

TM2 Group, LLC; New York (NY); On-site; USD 238,000 - 267,000; Full time

Benefits offered by this job

Competitive salary and comprehensive +

Generous vacation and holiday time off

Ongoing learning & development

Job summary

Protegrity is seeking a Senior Software Engineer, Ontology & Reasoning Systems to design and implement the ontology and reasoning layer that connects structured, semi-structured, and unstructured data into a single, explainable model for deterministic machine reasoning.

You will formalize complex domains, enabling traceable conclusions from facts, and use LLMs at the ingestion boundary for extraction, normalization, and categorization.

Qualifications

• 5+ years of backend experience in a modern language
• Strong foundation in logic and graph theory
• Ability to formalize domains as rules and constraints
• Experience with declarative or rule-based reasoning systems
• Experience building LLM-powered systems with validation workflows
• Knowledge graphs, RDF/SPARQL, and graph databases like Neo4j
• Provenability of outputs traceable to inputs

Responsibilities

• Own the ontology and rule layer unifying structured data into a single reasoning model
• Model entities, relationships, events, and constraints; derive conclusions
• Use LLMs at ingestion boundary for extraction, normalization, and categorization
• Ensure findings are traceable to source facts and rules
• Build retrieval capabilities (vector, keyword, graph traversal)
• Improve quality via entity resolution and constraint modeling
• Ship production services and research prototypes with correctness and reliability

Skills

Python

Go

Java

Scala

Rust

Discrete mathematics

Graph theory

Declarative reasoning

LLM pipelines

Knowledge graphs

Provenance

Tools

Neo4j

RDF

SPARQL

OWL

Job description

Protegrity is seeking a Senior Software Engineer, Ontology & Reasoning Systems to design and implement the ontology and reasoning layer that connects structured, semi-structured, and unstructured data into a single, explainable model for deterministic machine reasoning.

You will formalize complex domains, enabling traceable conclusions from facts, and use LLMs at the ingestion boundary for extraction, normalization, and categorization.