Build Notes

Project stories and technical notes

Multi-agent systemsAI evaluationProduction AIResearch methodsFounder engineering

Reader-first notesTechnical depth, clear boundariesFiltering: Research Methods

Peter McCann Strain working at a laptop on a sunny Glasgow balcony — From the writing desk

All AI Evaluation Computer Vision Founder Engineering LLM Cost Optimization Legal Infrastructure Medical AI Multi-Agent Systems Production AI Recruitment AI Research Methods

The Retrieval Bottleneck: A Controlled Comparison of 13 Deep-Research Architectures

AI Evaluation • Research Methods

Start here

Start with this

The Retrieval Bottleneck: A Controlled Comparison of 13 Deep-Research Architectures

13 min read

A controlled benchmark of 13 deep-research agent architectures on one shared tool layer — what happens to orchestration claims when you fix the model, the tools, and the judging protocol, and read the statistics honestly.

18 May 2026

Enter article

Guided entry

From Whole-Slide Images To Biological Structure: My Oxford DPhil In Medical AI

Medical AI • Computer Vision

10 min

From Whole-Slide Images To Biological Structure: My Oxford DPhil In Medical AI

A technical account of my Oxford DPhil: 3 million unlabelled cell nuclei from mouse placental histology, an 89% supervised classifier, near-perfect clustering of 8 cell types, and a self-supervised experiment that fell short of its goal.

20 May 2026

Satellite Pose Estimation and the Sim-to-Real Domain Gap

Computer Vision • Research Methods

14 min

Satellite Pose Estimation and the Sim-to-Real Domain Gap

How I approached spacecraft pose estimation at Astroscale: measuring the sim-to-real gap before modelling it, a preprocessing result that flipped sign with placement, the SPS-B architecture proposal, and AstroGAN — an image-translation model that closed the appearance gap while leaving the harder half open.

20 January 2026