AI systems researcher and Lead Gen AI Engineer building governed agents, temporal trust, and efficient edge AI. I work at Insight Global, contribute to DeepStation as an AI Advocate, and build open-source tools, research artifacts, and hands-on teaching materials.
Portfolio · Research · Software · Talks and workshops
| Project | Focus |
|---|---|
| Pollard | Governed execution trees, budgets, gates, and replay for AI agents. |
| Chronofy | Temporal validity and evidence decay for time-aware AI systems. |
| JSON-LD Ex | Machine-readable confidence, provenance, temporal scope, and interoperable knowledge. |
| Explainiverse | Explainability methods and evaluation with explicit audit scopes. |
| MultiSpecQR | Multi-layer QR encoding and neural-assisted decoding, with models and datasets. |
| Corpusgen | Speech-corpus design and evaluation for phoneme coverage across languages. |
Pollard's newer alpha companions are pollard-jev, for typed decisions and bounded simulated actions, and epicormic, for provider-drift detection and admission controls. See each project's documentation for its implementation and validation boundaries.
My research spans trustworthy AI, agent infrastructure, edge intelligence, semantic data, and efficient machine learning. The publication catalog distinguishes published contributions, accepted work, and preprints.
I teach through public workshops and runnable examples. Bringing the Heat includes slides and a Colab exercise for a live Hugging Face pipeline covering fine-tuning, evaluation, export, and inference. More materials are in the speaking archive.
PyPI · Hugging Face · ORCID · LinkedIn · Email