Research
The research behind the ideas
Write-ups hosted here and articles published on dermot-obrien.com, newest first. Each names the idea it belongs to; articles about the way the work is done name none.
How the agents are tested
The Earnings Lab has two agents, one that reads earnings releases into a graph and one that answers questions over it. This covers how each is checked, what happens to a failure, and what is not tested yet.
How the Earnings Lab is built
The Earnings Lab turns US quarterly earnings releases into an ontology-typed graph, answers questions over it as the signed-in user, and scores itself against the figures companies file in XBRL. This covers its parts on AWS and what it costs to keep and to run.
How the historian benchmark is built
A pre-registered benchmark that scores fault detectors on data kept the way a process historian keeps it, from public fault datasets through a simulated historian, four ways of reading the archive and a shared false-alert budget to decision rules frozen before scoring.
Ontology-guided GraphRAG
Question answering over a knowledge graph whose only schema is one curated OWL ontology, which shapes what gets extracted, how the graph is typed and how an agent queries it.
Watching historian data by its shape
Utilities already record years of readings from every transformer, pump and meter. We are testing whether reading that record by its shape rather than its value can warn of some slow faults sooner than an alarm limit does, recognise a fault from other assets in the fleet, and say what is likely to happen next.
Reading an irregular series as a few turning points
Real data arrives when events happen, not on a clock. Before looking for patterns, it helps to ask how few points are enough to keep the shape of a series.
Use AI to assess AI impact on your business
An LLM-scored impact assessment over a capability model, drawn as an interactive treemap.
dermot-obrien.com ↗
AI writes faster than your architecture team can stay consistent
Why architecture needs a typed, validated model once AI is writing the documents, and what that looks like.
dermot-obrien.com ↗
The researcher's fog of discovery (and why AI needs a lineage of ideas)
Why AI agents make an unmapped research process worse, and the lineage of ideas that lets them carry the load instead.
dermot-obrien.com ↗
Graph iterated function systems: adding structure to self-similarity
A mathematical bridge from fractal geometry to hierarchical time series and the embedding spaces that AI models operate on.
dermot-obrien.com ↗
Iterated function systems: the simple rules behind complex fractals
The theory from the Hutchinson operator to the Collage Theorem, and the first in a series extending it to hierarchical time series.
dermot-obrien.com ↗
Also open to you