Defensive Security AI Scientist

World Wide Technology Healthcare Solutions•Jenks, OK

7d•Onsite

About The Position

The emphasis of this role is to be the dedicated technical lead for developing WWT’s defensive AI operations, specifically focused on utilizing massive open-source models to solve the vulnerability management problem at an enterprise scale. This position requires an elite technical leader who is adept at advanced machine learning, can speak to complex exploit mechanics, and ensure our defensive posture rapidly evolves to keep pace with AI-accelerated offense and frontier-model-class intelligence. This role will also involve guiding the organization away from manual triage by anticipating exploit chaining and integrating the largest open-source AI models to ensure our defensive strategies remain ahead of advanced adversaries.

Requirements

Master’s degree or Ph.D. in Computer Science, Data Science, Artificial Intelligence, Cybersecurity, or a related highly technical discipline.
8+ years of progressive experience in data science, machine learning, or AI engineering, with a minimum of 3 years directly embedded in cybersecurity, threat intelligence, or vulnerability management.
Deep, hands-on expertise with the Hugging Face ecosystem (Transformers, Accelerate, PEFT).
Proven track record of deploying, quantizing, and fine-tuning (e.g., LoRA, QLoRA) large open-source models (Llama, Mistral, Qwen) for production environments.
Comprehensive understanding of the vulnerability lifecycle, patch gap analysis, exploit chaining, and standard industry frameworks (MITRE ATT&CK, CISA KEV, EPSS, CVSS).
You must understand how an adversary thinks to build models that stop them.
Expert-level coding proficiency in Python and leading deep learning frameworks, specifically PyTorch.
Extensive experience operating in high-performance computing environments.
You must be adept at managing GPU workloads, CUDA optimization, and deploying models on distributed infrastructure (e.g., Kubernetes, Ray, vLLM).
Strong capability in building high-throughput data pipelines to ingest, clean, and process massive datasets of threat intelligence, code repositories, and telemetry.
Demonstrated ability to own a complex problem space entirely, translating theoretical data science concepts into highly scalable, real-world defensive tools at an accelerated pace.

Responsibilities

Partner with the GS&A Cyber Practices to utilize to model and evaluate attack paths and how AI can help defenders operate at the speed of attackers.
Deliver measurable outcomes to customers and partners as to how AI and ML-based science improves cybersecurity defenses.
Participate in industry forums on the topic of using AI for defense and be able to perform technical demonstrations how AI is used for offensive operations.
Architect and deploy advanced machine learning models that analyze, score, and prioritize vulnerabilities at an unprecedented scale.
Continuously evaluate, integrate, and fine-tune frontier models to match and counter frontier-model-class attack methodologies and AI-driven exploit chaining.
Shift organizational operations entirely away from manual triage by providing automated, model-assisted workflows for immediate remediation guidance.
Ingest and model complex threat data, scanning code with open-source models to train systems on handling the exploding volume of automated vulnerability reports.
Dedicate 100% of operational bandwidth to defensive cybersecurity measures, using AI to neutralize emerging threats before they can be weaponized against our customers.
Act as the technical expert for data architecture, statistical modeling, and advanced machine learning applied to massive enterprise datasets.
Drive the vulnerability prioritization problem space autonomously from concept to customer-facing deployment without distraction.
Apply deep cybersecurity domain knowledge, specifically regarding vulnerability management, exploit mechanics, and patch gap mitigation, to all AI models.