AI Engineer — Multi-Agent LLM Security Orchestration

CipherData; Bellevue (WA); On-site; USD 120,000 - 155,000; Full time

Job summary

CipherData seeks an AI Engineer to design and operate the core multi-agent LLM system that analyzes telemetry from 70+ vendors. You’ll architect agent collaborations, orchestrate models for cost and latency trade-offs, and push research into production to support SOC teams.

You’ll build evaluation harnesses, measure output quality, and close the loop with data-driven improvements. This role emphasizes accuracy, reliability, and scalable deployment.

Qualifications

• Hands-on experience designing and building LLM applications (agents, RAG, orchestration, etc.).
• Strong software engineering fundamentals in Python.
• Experience building LLM workflows with various agent frameworks.
• Experience evaluating and improving LLM output quality with quantitative measures.
• Comfort defining ambiguous problems, experimenting quickly, and making data-driven decisions.

Responsibilities

• Build our multi-agent LLM system and improve how agents collaborate to detect threats.
• Create orchestration pipelines that balance cost, latency, and accuracy.
• Develop evaluation systems that measure output quality and feed learning signals.
• Experiment with prompting, architecture, and verification to reduce hallucinations.
• Contribute to next steps, including automated response and broader threat correlation.

Skills

LLM applications design

Python programming

LLM workflows

Evaluation & measurement

Data-driven decisions

Tools

Agent frameworks

LLM tooling

Job description

CipherData seeks an AI Engineer to design and operate the core multi-agent LLM system that analyzes telemetry from 70+ vendors. You’ll architect agent collaborations, orchestrate models for cost and latency trade-offs, and push research into production to support SOC teams.

You’ll build evaluation harnesses, measure output quality, and close the loop with data-driven improvements. This role emphasizes accuracy, reliability, and scalable deployment.