AI/ML practitioner with 10+ years building and deploying Generative AI and machine learning systems from prototype to production. I specialise in the full lifecycle of AI/ML: architecture, deployment, security hardening, and extracting measurable business value at scale. Extensive work with critical and regulated industries.
Currently working as a Staff-level AI Engineer in the Australian federal government, leading enterprise AI/ML strategy and production deployment. I oversee LLM infrastructure, workflow automation, and AI capability programmes.
Previously led high-impact production ML delivery across government and large enterprise (large-scale fraud detection capability recovering more than $2B+ annually) and served as Principal Consultant at a specialist consulting firm, managing enterprise AI engagements for clients including large banks, international government bodies, critical infrastructure, national security, and service delivery.
- Enterprise GenAI applications — OpenAI API integrations, production prompt engineering, evaluation harnesses, customer-facing AI products
- Agentic AI systems — multi-agent orchestration, tool use, code agents, automated reasoning pipelines
- RAG pipelines — embedding generation, vector search, semantic retrieval at scale
- Deployment infrastructure — secure inference, mTLS, ABAC, enterprise guardrails, streaming
- Workflow automation — n8n, Python, comparative low-code/pro-code architectures
- ML at production scale — NLP, computer vision, fraud detection, identity resolution
n8n-io/n8n — open PR #27309: adds mTLS client certificate support to the OpenAI node family (+675 lines, 27 new test cases). Covers OpenAI Enterprise mTLS beta, self-hosted inference endpoints behind certificate-authenticated proxies, and custom CA bundles. Addresses a recurring community pain point for enterprise deployments.
anthropics/anthropic-sdk-python — open PR #1280: adds a worked example (examples/mtls.py) demonstrating mTLS client certificate authentication and custom CA bundle configuration with the Anthropic SDK. No SDK code changes required — documents existing http_client capability that was previously undiscovered by enterprise users hitting proxy and certificate errors.
huggingface/smolagents (22k+ stars) — contributor. PR #1420: added reset_agent_memory support in GradioUI, merged by the HuggingFace core team.
n8n-nodes-mtls-openai — published npm package providing n8n community nodes for mTLS-authenticated connections to OpenAI-compatible inference endpoints. Solves a real enterprise gap: connecting workflow automation to LLMs behind certificate-authenticated proxies.
| Project | What it does |
|---|---|
| agentic-cti | Multi-agent threat intelligence system orchestrating specialised sub-agents (OpenCTI, OSINT, Wikipedia) — Hydra config management, automated evaluation harness, Docker deployment |
| alcohol-and-other-drugs-understanding | End-to-end RAG pipeline over medical literature — automated scraping, PGVector storage, LLM-powered chatbot via OpenAI-compatible API, UMAP + topic modelling, interactive data map |
| n8n-nodes-mtls-openai | Published npm package: custom n8n nodes (TypeScript) providing mTLS-authenticated connections to OpenAI-compatible inference endpoints — chat model and embeddings sub-nodes for enterprise RAG pipelines |
| hardened-llm-deployment | Production deployment blueprint: ABAC + mTLS secured LLM inference on Kubernetes/KServe/Istio — streaming responses with pre-computed classification, async audit logging |
| bellingcat-py | OSINT corpus analysis pipeline — entity relationship extraction, vector embeddings, agentic semantic search — paired with an n8n equivalent for direct pro-code vs low-code architectural comparison |
GenAI / LLMs: OpenAI API · HuggingFace · smolagents · RAG architectures · prompt engineering · ONNX · PyTorch Languages: Python (advanced) · TypeScript · SQL · Bash Infrastructure: AWS · GCP · Azure · Docker · Kubernetes · PostgreSQL · PGVector · Ray AI security: MITRE ATLAS · adversarial ML · differential privacy · ML red teaming
M.S. Analytics (Computational) — Georgia Institute of Technology, 2020 (GPA 4.0) B.Finance / B.Economics — Australian National University, 2012
Certifications: SANS Offensive AI (2025) · Stanford Professional Certificate in AI (2023) · GIAC Penetration Tester (GPEN) · GIAC Detection Analyst (GCDA) · Cloud Native Architecture (Udacity)