Research

AI & AI-integrated autonomous software.

I research how to integrate AI into production software so systems can decide, adapt, and act — not just report. My background in shipping business software means I think about this from the deployment end as much as the model end.

Focus areas

01
AI-integrated autonomous software How to embed predictive and decision-making AI into production business software so systems can act with appropriate autonomy, not just report.
02
Practical ML deployment in .NET environments Where ML.NET fits, where a Python service is the right answer, and where LLM APIs change the equation.

Current work

01
Agent architectures for IoT operations Extending EMOS and IWMS-style systems with agents that close the loop between prediction and action.
02
Observability for ML in production Drift detection, calibration monitoring, and audit trails for ML decisions inside business software.

Reading list

01
Designing Data-Intensive Applications — Martin Kleppmann Foundational for thinking about data flow in systems that mix transactional and ML workloads.
02
Reliable Machine Learning — Cathy Chen et al. Operational side of ML systems — directly relevant to AI integration in production software.

Collaborate

Open to research collaboration, particularly with industry teams looking at AI integration in operational software, or academics working on agent architectures for IoT and business systems. Get in touch.