CORTICAL
STACK_
Cyberware storage unit for experiments and field notes, pushing boundaries at the intersection of human creativity, machine intelligence, and agentic systems.
OPERATIONS
Field-deployed apps, systems and platforms at scale.
Agentic Engineering Coaching
Field-deploying agentic engineering practice with software teams. Disciplined coding-agent workflow staged as research -> plan -> implement -> validate, with context-window hygiene to fight context rot. Productized in a custom Claude Code and GitHub Copilot CLI template that orchestrates the stages through skills, label-driven state, hooks, and a Ralph-style autonomous harness that closes the issue-to-PR loop hands-free.
DOSSIER
Field-deployed AI Engineer for over 6 years. Previous tours included Lufthansa Group, leading the One Data Platform AI squad and driving ML, MLOps, and GenAI engineering across the Group.
Earlier postings: SAP HANA technical architect at Nestlé, then engineered AC Immune's first cloud-hosted Azure ML platform. Foundations laid in data engineering, ML, and applied research before the GenAI wave.
Master of Intelligent Systems with distinction from De Montfort University, 2021. 90% average across modules and thesis. Off-grid: two beautiful daughters, kiteboarding, a book a week.
TECH
Tools of the trade.
AI / ML
- Python
- LangGraph
- Agents
- RAG
- Fine-Tuning
- LLMs
- Foundry SDK
- MLOps
- GenAIOps
- Prompt Engineering
- Agentic Engineering
- Responsible AI
- LLM-As-A-Judge
- Multi-Modal
Cloud
- Azure
- Bicep
- Terraform
- APIM
- Docker
- CI/CD
- Security
Data
- Ontologies
- OWL 2
- Knowledge Graphs
- SPARQL
- Azure ML
- Vector Search
- SQL
- Pandas
Tooling
- Claude Code
- Copilot CLI
- Bash
- Jupyter Notebooks
- VS Code
- Git
- GitHub Actions
ACADEMIC
ARCHIVES
9 declassified case files from the Master of Intelligent Systems program. De Montfort University, 2019-2021. Graduated with distinction.
Masters Thesis: COVID-LEAP
Azure-hosted knowledge base with BERT semantic search across 4.7M coronavirus research paragraphs and 500K clinical trials.Applied Computational Intelligence
Predicting customer churn, appetency, and up-selling for KDD Cup 2009. Gradient boosting placed 34/89.Computational Intelligence Optimisation
HOP: Python hyper/meta-heuristic platform (SA, GA, PSO, DEA, ES) on the flow-shop scheduling problem and continuous benchmarks.Artificial Neural Networks
Neural-network intrusion detection on the DARPA/MIT Lincoln dataset, plus NDVI-based crop prediction with ANNs.Artificial Intelligence Programming
Natural-language movie-database querying with NLTK + WordNet generating SQL, plus a Prolog/NLP state-of-the-art review.Data Mining
Cardiovascular-mortality risk classifiers on the Framingham Heart Study (SAS Enterprise Miner): decision trees, neural networks, gradient boosting.Fuzzy Logic
Python fuzzy inference system for breast-cancer tumour classification on the Wisconsin dataset. Beat neural nets, decision trees, and random forests.Mobile Robots
V-REP simulated robot with ultrasonic sensors learning to wall-follow. PI/PID controller cut navigation error 55.8%; SAR variant with A* path-finding.Research Methods
Critical evaluation of an immersive-VR analytics paper, and a PhD proposal for augmented analytic knowledge discovery.TRANSMISSIONS
Recent operational dispatches from the field on machine learning, GenAI, and software engineering.
Peter Steinberger 1 month token spend
1 MIN$1.3M!
Awesome Foundry Nextgen
1 MINAwesome Foundry Nextgen
LLM wiki ops
2 MINllm wiki ops
AI Tour 2026 Zurich
1 MINFoundry Control Plane booth