AI Safety Argumentation Platform Engineer

Aimlroles; Northern (KY); Hybrid; USD 110,000 - 170,000; Full time

Job summary

CARMA seeks a researcher to build and operate a system that makes the epistemic backbone of AI risk management both formally tractable and practically persuasive. You will extend ontologies and knowledge graphs, implement defeasible argumentation, and run LLM-assisted pipelines with expert curation.

With a focus on safety, governance, and frontier AI dynamics, you will map strongest counterarguments, engineer agent coordination patterns, and ensure outputs meet diverse communications needs for

Qualifications

• Must have formal or semi-formal argumentation theory knowledge.
• Experience with ontology or knowledge graph development is required.
• Operational experience with LLM agent systems and QC regimes is essential.
• Proven ability to prototype and ship in production-adjacent contexts.
• Substantive grounding in AI safety, governance, and frontier AI dynamics.
• Familiarity with evidence in philosophy of science and related concepts.
• Strong coding skills and comfort with graph databases or query languages.
• Experience designing cross-check and verification scaffolds for automated processes.
• Good judgment on well-supported claims vs hedging or withdrawal.
• Self-directed and able to communicate in writing.

Responsibilities

• Extend ontologies and knowledge graph schemas for claims, evidence, and argument structures.
• Implement defeasible argumentation frameworks capturing structure and rebuttal vulnerability.
• Operate LLM-driven population pipelines with cross-checks and human-in-the-loop curation.
• Architect multi-step research task patterns with robust error handling.
• Pre-harden argument structures by mapping counterarguments and defeaters.
• Build export pipelines translating argumentation into diverse communications formats.
• Maintain awareness of AI safety and governance developments for graph updates.
• Collaborate with communications staff and researchers to meet persuasive needs.

Skills

Argumentation theory

Ontology engineering

LLM agent systems

Vibecoding practice

AI safety & governance

Philosophy of science

Graph databases

Cross-check scaffolds

Judgment & hedging

Written communication

Education

Graduate work in argumentation theory or related field

Tools

OWL/RDF

Graph databases

Query languages

Job description

CARMA seeks a researcher to build and operate a system that makes the epistemic backbone of AI risk management both formally tractable and practically persuasive. You will extend ontologies and knowledge graphs, implement defeasible argumentation, and run LLM-assisted pipelines with expert curation.

With a focus on safety, governance, and frontier AI dynamics, you will map strongest counterarguments, engineer agent coordination patterns, and ensure outputs meet diverse communications needs for