ADVENTURES IN

All our insights across AI, technology, and innovation.

Showing 16 posts

The Missing Piece of the Claude Code Workflow: Isolated Worktree Databases

Git worktrees + Claude Code need isolated databases. Here's how to spin up separate Supabase stacks per branch with automatic port allocation.

#claude-code#git-worktrees#supabase#local-development#developer-productivity

Duncan Anderson·Feb 2, 2026

OpenClaw and the Programmable Soul

Four primitives. Agent societies. And a preview of what enterprise AI might become.

#AI Agents

Thibault Boutet·Feb 2, 2026

Building AI products means managing API costs. Here's what actually worked.

A practical guide to making LLM costs debuggable in production: how to log tokens at the right granularity, understand prompt caching behavior, and identify the real drivers behind cost spikes.

#AI Agents

Duncan Anderson·Jan 19, 2026

Seatbelts for AI: Lessons from the Grok Image Controversy

Grok's image abuse problem is engineering, not ideology—and we already know how to fix it

#AI Ethics

Thibault Boutet·Jan 16, 2026

Building a Production Multi-Agent System for Document Writing

A deep dive into the architecture of a briefing generation system that uses 15 specialized AI agents to transform PDF source materials into structured documents.

#AI Agents

Kevish Napal·Dec 18, 2025

Mean Pooling Beats Attention: Predicting Telomerase Activity from Whole-Slide Images

From global signals to local clues with ABMIL—and what it taught us about telomerase as a tissue-wide phenotype.

#Machine Learning

William Auroux·Nov 19, 2025

The Dimension Dilemma: Why 2.5D Models Outperform 3D CNNs for Stroke Classification

Lessons & Experiments on training deep learning models on 3D medical data.

#Machine Learning

William Auroux·Nov 12, 2025

Understanding and Processing CT Imaging for Stroke Detection

A practical guide to turning raw brain CT into training-ready data.

#Machine Learning

Kevish Napal·Nov 7, 2025

Building a Reproducible Multimodal Pipeline

Part-2 in a series of posts about proteomics, cancer biology, and AI-driven solutions for oncology.

#Machine Learning

Kevish Napal·Nov 1, 2025

Understanding Cancer and Telomerase: From Biology to New Treatments

Part-1 in a series of posts about proteomics, cancer biology, and AI-driven solutions for oncology.

#Machine Learning

Duncan Anderson·Oct 4, 2025

Stop Treating AI Like Magic: Why Context Engineering Beats Bigger Models

Smaller models with a disciplined context rival flagships at a fraction of the cost

#Context Engineering

Duncan Anderson·Sep 23, 2025

The 4 Levels of AI Agents: When to Use Workflows vs Autonomous Systems

Stop over-engineering simple problems and under-engineering complex ones

#AI Agents

Duncan Anderson·Jul 23, 2025

Barnacle Labs Appoints Dr. Oliver Bogler as Scientific Advisor

Distinguished Cancer Researcher and Federal Program Leader to Accelerate AI-Driven Scientific Discovery

#Company

Duncan Anderson·Jul 15, 2025

Why We're Going All-In on Forward Deployed Engineers for AI Projects

Forward Deployed Engineers help to scope and define new projects

#Company

Duncan Anderson·Jan 20, 2025

Me and an AI: Building a New Website Together

Our new website was built by Claude and o1. A human (me) oversaw its construction, but AI wrote every line of code.

#Company

Duncan Anderson·Feb 3, 2024

Beyond Data Hoovering: The Nuanced Reality of Training Large Language Models (LLMs)

Training Large Language Models (LLMs) is an evolving science — or, perhaps, an art form. In this post I set out to shed some light on exactly what is meant by training a model.

#AI-Training