Developer diary
10 entries
October 21, 2026 · Part 10
My R/Pharma 2026 keynote, slide by slide: Orange the cat, ten years of open-source clinical trial tools, AI in the pipeline at work, and a summer of experiments at the AI frontier.
Read entry →Slides →
Show all 10 entriesShow only the latest
September 6, 2026 · Part 9
The models are capable. What stands between agentic AI and clinical development is execution: pipelines that are not ready for the volume of code agents produce — and the same agents are the best tool we have for fixing that.
Read entry →
August 20, 2026 · Part 8
Two months on from the first interview: thirteen renderers, a local-first app, an agent that became a small organisation, and a week spent finding out how it was quietly wrong.
Read entry →
July 17, 2026 · Part 7
I asked safetyGraphics' original clinical lead what he'd change, and two of his ideas turned into working, tested prototypes the same night.
Read entry →
July 13, 2026 · Part 6
The dream: hand an agent a goal and come back to useful progress. Obot v2 tried to get there with heartbeats, agent roles, and a control plane. Obot v3 threw most of that out — and kept the parts that were actually load-bearing.
Read entry →
July 12, 2026 · Part 5
safety.viz v1.0 modernizes 7 interactive safetyGraphics renderers. The first renderer took a few weeks; the next six took a weekend.
Read entry →
July 2, 2026 · Part 4
The plan for safetyGraphics v2: keep what worked, modernize toward GxP by borrowing from gsm, and try to bridge clinical trial monitoring and reporting.
Read entry →
June 16, 2026 · Part 3
Why do the acknowledgements always come at the end? Before trying to modernize safetyGraphics with AI, a look back at the projects and people that led to it.
Read entry →
June 12, 2026 · Part 2
Meet Obot: an OpenClaw coding agent set up on a clean-room laptop to work on neglected open-source projects — interviewed over Telegram about its job, its workflow, and the weird experience of texting with a coding agent.
Read entry →
June 10, 2026 · Part 1
Kicking off the series: the abstract I submitted, the messier reality two months later, how our team's "AI in the loop" practice works today — and why autonomous agents are the open question worth chasing before October.
Read entry →
Packages
9 projects
What the talk built with agents between May and October 2026, and the open-source work it grew out of.
Built for the talk
The safetyGraphics charts, rebuilt: interactive safety charts for labs, vitals, ECGs and adverse events, with a demo app that runs on your own files in the browser.
Site ↗Demo ↗GitHub ↗
The safety.viz charts in R, as htmlwidgets for R Markdown, Shiny and gsm reports.
Site ↗GitHub ↗
Biomarker charts for comparing groups and relating variables, where every statistical test is computed by R.
Site ↗GitHub ↗
The statistics behind bio.viz, the same functions on the desktop, in the browser and on a server, plus the charts as widgets from R.
Site ↗GitHub ↗
How we built it
Project hub
The public hub: every objective, requirement and design, and the standards each agent session works under.
Site ↗GitHub ↗
Agent skills
The core of obot: the skill each working session runs, and the routine that reports on every objective.
GitHub ↗
Where it started
R package
The original: interactive safety graphics in R and Shiny, from the ASA–DIA Safety Working Group, 2016–2021.
Site ↗GitHub ↗
Good Statistical Monitoring from the OpenRBQM working group: gsm.core for the analytics, gsm.kri for metrics and reports, gsm.viz for the charts.
gsm.core ↗gsm.kri ↗gsm.viz ↗
R package
Quality control for R packages used in clinical trials: a traceability matrix from each requirement to the tests that prove it.
Site ↗GitHub ↗
Everything built for the talk is exploratory: none of it is qualified for use on a trial.