Skip to content

R&D

Research

Not published papers. These are the areas where we have built enough to have formed views, including views that changed.

What this is, and what it is not

We are an engineering company, not a research lab. We have not published papers, we do not claim novel contributions, and we are not going to describe standard practice as innovation.

What we do have is accumulated judgement from building these systems — the decisions we now make by default, the ones we changed our minds about, and the questions we still cannot answer well.

Areas

Agent systems

Scope, tool access, handover, and what makes an agent safe enough to give real access.

LLM engineering

Why retrieval quality dominates model choice, and what evaluation actually requires.

Applied automation

Which steps should stay with a person, and how failure should behave.

Start here

Tell us what is slow, manual, or breaking.

Answer a few questions and get a written brief back — scope, proposed architecture, and what it would take to build.

Start a projectTalk to someone