Skip to content
All jobs

Senior AI/ML Engineer

Dragos

United States $190K – $190K Posted 6d ago

About the role

At Dragos, the mission is personal. The systems we protect deliver the water you drink, power your home, and keep the hospitals your community depends on running. Those critical infrastructure systems that power our civilization around the world are under attack every day by adversaries. When those systems fail, people are

immediately

at risk.

We are the global leader in

xOT

cybersecurity, combining technology, threat intelligence, and expert services. The people here chose this work because they understand

what is

at stake

. Here, you will find a remote-first mission-driven team across North America, Europe, the Middle East, and APAC built on authenticity, transparency, and trust. If safeguarding the systems that protect your family, friends, and community is the kind of work that matters to you, you are in the right place.

About the Role

:

We’re looking for a Senior

AI/ML Engineer

to join our Engineering team. You’ll take existing large language models and embedding models and put them to work in our product and data pipelines. You won’t be training foundation models from scratch. You will be doing the careful applied work: deciding which models and techniques fit which problems, building the harnesses that wire them into our workflows, proving through evals that the outputs are reliable, and shipping those solutions to production on infrastructure you define as code.

You’ll work closely with AI Engineers, Data Engineers, and product teams to bring LLM-driven features into the Dragos platform. Examples include summarizing and triaging network activity, enriching and classifying assets, retrieving relevant threat intelligence through semantic search, and helping ICS/OT security analysts focus on what matters.

Responsibilities

:

Apply LLMs and embedding models to cybersecurity data problems in the ICS/OT domain, including retrieval-augmented generation, semantic search, extraction, summarization, and classification.

Design and build LLM harnesses: prompt and context management, tool and function calling, agent loops, structured outputs, retries, and guardrails.

Build and maintain embedding pipelines, including chunking strategies, vector storage, retrieval tuning, and re-ranking.

Design and run LLM evaluations: build golden datasets, define task-specific metrics, use LLM-as-judge where it fits, and catch regressions when models, prompts, or data change.

Integrate LLM outputs into existing data pipelines and product workflows, supporting both batch and near-real-time processing patterns.

Deploy and operate your solutions with Terraform and Kubernetes. Define infrastructure as code, package services in containers, and manage releases across environments.

Work with Data Engineers to give LLM-driven pipeline stages clear data contracts, good observability (latency, cost, token usage, output quality), and sane failure modes.

Evaluate open-source, self-hosted, and commercial models against specific use cases. Know when to use prompting, retrieval, or tool use, and when to escalate to fine-tuning or a model-building effort.

Write clean, maintainable Python or Rust that other engineers can reason about, test, and extend.

Troubleshoot LLM behavior in production, diagnosing hallucinations, retrieval misses, prompt regressions, drift after model upgrades, and unexpected edge cases.

Communicate clearly about what a model is doing, where it’s uncertain, and how its outputs should and shouldn’t be used downstream.

Qualifications

:

5+ years of software engineering experience, including meaningful time building with LLMs or running data pipelines in production.

Strong Python skills, SQL proficiency, and comfort reading and reasoning about data at scale.

Hands-on experience building LLM applications, such as RAG systems, tool-using agents, or structured extraction pipelines, using commercial APIs or self-hosted models.

Practical experience with embeddings and vector search: choosing embedding models, chunking, indexing, and measuring retrieval quality.

Experience designing and running LLM evals, and the judgment to tell whether outputs are trustworthy for a given use case, beyond whether a benchmark looks good.

Hands-on experience deploying production services with Terraform and Kubernetes, including Helm or similar tooling, CI/CD pipelines, and secrets and configuration management.

Solid understanding of data pipeline concepts: how data flows, where it gets transformed, what can go wrong, and how to make failures visible.

Strong written and verbal communication, and ability to explain tradeoffs to technical and non-technical stakeholders.

Cybersecurity domain knowledge is a meaningful plus but not required, especially in threat detection, network behavior, or ICS/OT operations.

Nice to Have

Experience with the Model Context Protocol (MCP) or building tools and integrations for LLM agents.

Experience self-hosting and serving open-weight models (vLLM or Ollama), including GPU scheduling on Kubernetes.

Experience with hybrid search (keyword plus vector) or with graph-based representations of network topology or asset relationships.

Familiarity with stream processing or event-driven architectures, such as RabbitMQ or Kafka.

Experience with LLM observability and tracing tools, such as OpenTelemetry or Langfuse.

Compensation

:

Salary:

$190,000

Competitive Equity Package

Comprehensive Benefits Plan

#LI-NH1 #LI-REMOTE

Dragos is an Equal Opportunity Employer and considers applicants for employment without regard to race, color, religion, sex, orientation, national origin, age, disability, genetics, or any other basis forbidden under federal, state, or local laws. All new hires must pass a background check as a condition of employment.

    Cookies on Rolebound

    We use necessary cookies to sign you in and keep the site working — those cannot be turned off. We would also like to set analytics and advertising cookies so we can see which pages help people get interviews and which adverts brought them here. Nothing optional is set unless you say yes, and you can change your mind at any time from “Cookie settings” in the footer. Read our Privacy Policy.