Machine Learning Engineer – Security Data Modeling

Company Description

Beacon is rebuilding the data and context layer for modern security operations. 

Today’s security teams are drowning in noisy, inconsistent telemetry collected through brittle pipelines and legacy tooling that were never designed for the speed, scale, or complexity of today’s environments. We’re a team of engineers, researchers, and builders creating a security-first data platform that shapes, enriches, and routes the right data to the right systems – from SIEMs to data lakes to AI-powered investigation tools. 

Backed by deep incident response and offensive security expertise, we’re helping organizations regain control of their telemetry, reduce operational drag, and unlock the visibility they need to defend at modern scale. We’re still early, growing fast, and looking for people who want to work on hard technical problems that materially change how defenders operate.

Role Overview

We’re looking for an ML engineer who understands both machine learning and the messy, heterogeneous nature of security telemetry.

What You’ll Do

• Build models for correlation, entity resolution, and enrichment.
• Collaborate with engineering to operationalize ML in real-time pipelines.
• Develop evaluation frameworks for model performance.
• Partner with research on attacker-informed features.
• Build scalable training and inference infrastructure.

What We’re Looking For

• 10+ years ML engineering experience.
• Strong fundamentals in data modeling and feature engineering.
• Experience with structured event data.
• Proficiency in Python + ML frameworks.

Nice-to-Haves

• Security or observability background.
• Graph modeling or time-series experience.
• Familiarity with vector DBs or LLM-based enrichment.

What Success Looks Like

• Models improve correlation, enrichment, and coherence.
• ML systems run reliably at scale.
• Product teams build on top of your ML primitives.

Why Beacon

At Beacon, you’ll work on problems that matter: building the data foundations that modern security operations – and the next generation of AI-driven defense – will rely on. 

We’re a low-ego, high-ownership team that values clear thinking, strong engineering fundamentals, and the ability to turn complex ideas into real systems used by real security teams. You’ll collaborate directly with experienced IR and offensive researchers, shape products with immediate customer impact, and operate in an environment where good ideas move quickly from design to production. 

If you want to work on deep technical challenges, contribute to an architecture-first approach to security data, and help define a new category from the ground up, you’ll feel at home here.