I still like getting into the work itself.
I use AI for research, product development, implementation and review. My projects explore practical questions about connecting systems, choosing models, controlling costs and checking the results. They keep me close to the engineering and operating decisions behind a software business.
Coding assistants · Infrastructure
Brainstorm
Tools for working with AI coding assistants and connecting them to infrastructure. The project brings together model routing, a command-line interface, MCP connectors, cost tracking and evaluation tooling.
The work includes approval workflows and audit reporting: giving an assistant useful access while keeping its actions understandable and controllable.
Explore the Brainstorm repository ↗
Security software · Development experiment
Learning from the build
Brainstorm Security Stack is an open experiment in building security software with AI agents. The work includes cloud resource scanning, policy evaluation, authentication, architecture documents and code reviews.
The development process is part of the project: what agents produce, where they fail and how review and testing change the result.
Explore the security project ↗
Product design · Shared foundations
A consistent product experience
The Brainstorm Design System provides shared typography, color, spacing, motion and interface states for related products. It gives the projects a common foundation as they develop.
Explore the design system ↗