AI Security Consulting
Security for teams building
with LLMs
Prompt injection, data leakage, agentic risk, supply chain exposure — these aren't caught by traditional security tooling. I work with engineering and security teams to find and fix them before they become incidents.
Security Assessment
A structured technical review of your LLM deployment. Findings ranked by severity, with reproduction steps and specific remediation guidance.
Learn more →Implementation Consulting
Embedded alongside your team during the build — designing security in from the start rather than retrofitting it after launch.
Learn more →Team Enablement
Hands-on workshops that give your engineers and security team shared vocabulary and practical skills for LLM-specific risks.
Learn more →Focus areas
What I assess and secure
Specific expertise in how LLM systems fail — not generic cybersecurity rebranded.
Prompt Injection
Direct and indirect injection, jailbreaks, and multi-turn manipulation that override system intent.
Data Leakage
Paths where training data, system prompts, or user data can be extracted through model outputs.
Agentic System Risk
Tool-using agents introduce privilege escalation and confused deputy attacks that don't exist in traditional systems.
Supply Chain & Model Risk
Third-party models, RAG data sources, fine-tuning datasets, and embedding pipelines as attack vectors.
Access Controls
How user identity, permissions, and data isolation are enforced — or not — across your LLM infrastructure.
Monitoring & Response
Logging, anomaly detection, and incident response playbooks built for how LLM systems actually fail.
I've spent years building and deploying LLM systems for enterprise clients — working across the full stack from model selection and fine-tuning to production infrastructure on Kubernetes. The gap I kept running into: engineering teams building sophisticated AI systems without security expertise embedded in the process, and security teams asked to review systems they didn't fully understand.
Molntek is how I work with organizations that need someone who can speak both languages fluently — deep enough in AI engineering to understand how these systems actually behave, and security-focused enough to find where they break.
Ready to talk?
Tell me what you're building. I'll tell you honestly whether I can help and what that would look like.