About
I think in products, and I build the systems behind them.
Product engineer and strategist. Founder of BlindspotLab and creator of Arcapush, the discovery and distribution registry for products, AI agents and hackathon builds, and Ìhìn, an AI-native creator distribution platform. Also behind the open-source developer tools StackBrief, StyleGraft and the Arcapush CLI.

- Story
My background is in history education. That training is still how I work: I start with context, ask what changed and why, and look for the explanation underneath the obvious one.
Today I turn rough ideas, useful information and overlooked problems into products people can use. I usually own the whole path: the strategy that decides what should exist, the architecture that lets it last, and the engineering that ships it.
- Now
I’m the founder of BlindspotLab (BlindspotLab Limited, RC 9855946, incorporated 14 September 2026) and the creator of Arcapush, which I’ve built since February 2026. I also maintain three open-source developer tools on npm. See what I’m working on now.
- Evolution
- Feb 2026
First commit on 16 Feb 2026. It has since become the discovery and distribution registry I maintain today.
- Apr 2026
A builder marketplace with USDC escrow on Base, and the GitHub-to-car experiment that later became Admon.
- Jul 2026
StackBrief v1 on npm (OpenAI Build Week), Revel's agent and payment stack, and Admon's contracts on Monad Mainnet.
- Aug 2026
StyleGraft and the Arcapush CLI
Two more open-source npm packages: design systems from visual references, and terminal and MCP submission for Arcapush.
- Sep 2026
BlindspotLab Limited incorporated
Incorporated in Nigeria on 14 Sep 2026 (RC 9855946). The same month: AITAX client engineering, Outship and Ratssacre.
- Oct 2026
Guided terminal onboarding and MCP configuration, released 8 Oct 2026.
- Feb 2026
- Principles
Understand before changing
Most bad changes come from an incomplete picture of the system. I trace the real architecture first; StackBrief exists because of this habit.
Improve before rebuilding
A rebuild brings new risks. I ask what must stay untouched and whether a focused fix solves the actual problem.
Evidence over adjectives
Numbers ship with their definitions. Products should be able to show what they did, not just claim it.
AI as leverage, not judgement
Your input shapes the output. I treat prompts, evals and quality gates as engineering, and I credit them as such.
- DAETO
The framework I use to move from an unclear idea to a product that can be built, shipped, measured and improved.
Discovery
The problem, the people, the context and the real constraints.
Approach
Direction, positioning, architecture and the execution path.
Execution
The smallest strong version, built with clear priorities and disciplined scope.
Tracking
What is working, what is failing, and what users actually do.
Optimisation
Improvement from evidence, not assumptions or random feature additions.
- Credentials
11 Anthropic Academy certificates. Each link opens its public verification page.
- Claude 101
- Claude Code 101
- Claude Code in Action
- Building with the Claude API
- Introduction to Model Context Protocol
- Model Context Protocol: Advanced Topics
- Introduction to Agent Skills
- Introduction to Subagents
- AI Fluency: Framework & Foundations
- AI Capabilities and Limitations
- Introduction to Claude Cowork
- Elsewhere