Rida Al Barazi · Pickering, Ontario · Greater Toronto Area
Software architect, founder, and product builder.
I build software where product, architecture, and operating reality collide. My current focus is AI systems that keep state, tools, permissions, evaluation, and human control explicit in the application.
I’m the founder and principal of Amidship, work on product and software architecture at Salesforce, and build Agentify, PDFSight, and BranchBox.
Current work
What I’m building and operating.
I use my own products and open-source work to test the same ideas I bring to architecture and client work.
- 01
Amidship ↗
A product and engineering studio for difficult software systems.
- 02
Agentify ↗
Application-owned state, tools, knowledge, and human control for AI workflows.
- 03
PDFSight ↗
A document workflow that uses AI to understand forms while deterministic software owns construction and delivery.
- 04
BranchBox ↗
Isolated development runtimes for parallel human and AI work.
Writing
What I’m thinking through.
Most posts start with something that broke, surprised me, or changed how I work.
· 10 min read
Why I finally tried OpenClaw (and how I'm making it safe)
I avoided running agents with real access for a year. The unlock wasn't better models or tighter sandboxes. It was giving the agent its own identity.
· 3 min read
MCP's Auth Problem
The MCP spec expects oauth at connect time, before you know what tools you need. Here's the chicken-and-egg problem I hit and the workaround I landed on.
· 9 min read
YOLO Mode Only Works When YOLO Can't Hurt You
What I learned from 12 months of coding with AI agents—and why I built isolated dev environments to make them actually useful.
Talks
Recent talks, in full.
Both 2026 presentations are available as live slides, with their source repositories and official ConFoo pages.
ConFoo 2026 · Montréal
Agentic Coding: Building Features with AI Teammates
A live case study in rebuilding ConFoo’s website with parallel coding agents, structured specifications, isolated branches, and automated cross-model review.
ConFoo 2026 · Montréal
Safe Agentic Dev Environments
Safe Agentic Coding — Environment Engineering for Reliable AI Teammates
A practical framework for giving coding agents enough access to test and debug their work while keeping environments, credentials, and verification paths bounded.
Background
A career across startups, products, and enterprise systems.
I’ve spent more than two decades moving between founder-led companies and large product organizations. That range is why I tend to work across product decisions, architecture, and implementation instead of treating them as separate roles.