The Role
Lead design, training, and deployment of small, efficient language models and agentic systems for security: threat detection, policy enforcement, real-time inference, edge deployment, adversarial robustness, and productionization in security pipelines.
Summary Generated by Built In
AI Lead (Small Language Models + Cybersecurity)
About the Role
What You’ll Do
What We’re Looking ForCore Requirements
Location: San Francisco Bay Area (preferred) / Remote
Company: Protocol Nine
We are looking for an AI Lead to build the core intelligence layer of an AI-native security platform. This role sits at the intersection of small language model (SLM) development, agentic systems, and cybersecurity, and will define how models reason about intent, risk, and behavior in real-world environments.
You will lead the design, training, and deployment of specialized models that operate in constrained, high-performance environments (e.g., edge, real-time inference), and apply them to security problems such as threat detection, policy enforcement, and autonomous decision-making.
Design & Train Small Language Models (SLMs)
- Build domain-specific models optimized for latency, cost, and controllability
- Fine-tune models on security datasets (logs, network traffic, code, policies)
- Develop techniques for distillation, quantization, and efficient inference
Build AI-Native Security Systems
- Architect models that reason about intent in security contexts
Develop detection systems for threats across:
- Network traffic
- Application behavior
- Integrate models into real-time decision pipelines (e.g., firewall, policy engine)
Agentic AI & Decision Systems
- Design multi-agent systems for continuous monitoring, analysis, and response
- Implement feedback loops between detection, reasoning, and enforcement layers
- Ensure reliability, explainability, and controllability of autonomous systems
Model Infrastructure & Deployment
- Optimize models for edge + distributed environments
- Build evaluation frameworks for adversarial robustness and false positives
- Work closely with engineering to productionize models (APIs, pipelines, scaling)
Security Research & Innovation
- Stay ahead of emerging threats (e.g., AI-generated attacks, supply chain risks)
- Experiment with novel approaches (e.g., semantic code analysis, intent verification)
- Contribute to technical strategy and product direction
5+ years in machine learning / AI engineering (or equivalent depth)
Hands-on experience training or fine-tuning small or specialized language models
Strong understanding of:
- Transformer architectures
- Model optimization (quantization, pruning, distillation)
- Evaluation and benchmarking
Experience in at least one area:
- Network security / firewalls
- Endpoint or cloud security
- Application security or code analysis
Familiarity with:
- Threat detection systems
- Logs, telemetry, and security data pipelines
- Adversarial attack vectors
- Strong programming skills (Python + ML frameworks like PyTorch/JAX)
- Experience deploying models in production environments
- Understanding of distributed systems and real-time inference constraints
Experience with edge AI or low-latency systems
- Familiarity with agent frameworks / multi-agent systems
- Contributions to open-source ML or security projects
- Competitive salary + equity
- Early-stage ownership and high impact
- Opportunity to define a new category in security
Skills Required
- 5+ years in machine learning / AI engineering (or equivalent depth)
- Hands-on experience training or fine-tuning small or specialized language models
- Strong understanding of transformer architectures
- Experience with model optimization techniques (quantization, pruning, distillation)
- Experience in at least one cybersecurity area: network security, firewalls, endpoint or cloud security, or application security/code analysis
- Familiarity with threat detection systems, logs, telemetry, and security data pipelines
- Understanding of adversarial attack vectors and adversarial robustness evaluation
- Strong programming skills (Python) and experience with ML frameworks like PyTorch or JAX
- Experience deploying models in production environments (APIs, pipelines, scaling)
- Understanding of distributed systems and real-time inference constraints, including edge and low-latency environments
- Experience with edge AI or low-latency systems
- Familiarity with agent frameworks / multi-agent systems
- Contributions to open-source ML or security projects
Am I A Good Fit?
Get Personalized Job Insights.
Our AI-powered fit analysis compares your resume with a job listing so you know if your skills & experience align.
Success! Refresh the page to see how your skills align with this role.
The Company
What We Do
Protocol Nine is a stealth-mode startup dedicated to redefining the network security space. The company is developing next-generation security infrastructure and an AI-native security platform. By leveraging specialized small language models and agentic systems, Protocol Nine aims to provide real-time threat detection and automated policy enforcement to protect critical assets and ensure robust security in evolving digital environments.
