Ainsley Woo — Custom AI Agents & AI Security
Cybersecurity & AI Researcher · Singapore · SUTD
I'm Ainsley Woo, a cybersecurity & AI researcher in Singapore (SUTD, A*STAR Labs). I build custom AI agents wired into your workflow, fine-tune and train custom AI models, and research how intelligent systems fail — LLM and agent security, adversarial AI, and AI red-teaming. Available for hire to scope, build, evaluate and deploy bespoke AI agents.
Work with me — custom AI agents, models & security
- Custom AI agents — from S$3,000
Bespoke agents wired into your workflow. - Agent maintenance & monitoring — from S$450/mo
I keep your deployed agents reliable — you bring the API keys. - Custom AI models · fine-tune — from S$4,500
Fine-tune an existing model on your data. - Custom AI models · from scratch — from S$12,000 + compute
Your own architecture, built and trained from the ground up. - Full-stack build, deploy & maintain — build from S$2,500 · maintenance from S$300/mo
Sites and agents shipped end to end. - Pentesting — from S$2,500
Security testing for apps and AI systems. - Design & planning (consult) — from S$1,000
Scoping, architecture, and a written plan before the build — credited toward the project if you proceed.
Selected work
- SentinelLM (2026) — A runtime adversarial firewall for LLM applications.
- AgentGuard (2026) — Dual-stream anomaly detection for LLM-agent servers.
- BenchAudit (2026) — A meta-evaluator that audits agentic-AI benchmarks.
- AI Attack Validation Suite (2026) — Complete coverage of 92 AI/LLM attack types.
- Zero-Trust Attack-Path Mapper (2026) — Maps identity and lateral-movement paths across clouds.
- Malware Behavior Attribution (2026) — Attributes malware from behavior, not signatures.