Thoughts & Writings
Ideas pinned to the board — tap a cluster to fan it out.
pinned thoughts, published and in progress
Seshan: What 280 Hours of Building Taught Me About AI, Learning, and Letting Go of Perfection
Seshan: What 280 Hours of Building Taught Me About AI, Learning, and Letting Go of Perfection
A raw, honest account of building an AI-powered platform — the wins, the failures, and what perfection actually costs.
Five AI Agents, One Portfolio: The Architecture Behind Automation
Five AI Agents, One Portfolio: The Architecture Behind Automation
How Scout, Ticker, Matcher, Quill, and Sage were built — and why the human never left the loop.
Vibe Coding Is Real — what happens when a non-developer builds a full stack app
Vibe Coding Is Real: What Happens When a Non-Developer Builds a Full Stack App
My honest account of building suniliyer.ca — without being a traditional developer.
Why I built five agents instead of one
Why I Built Five Agents Instead of One: The Case Against the Swiss Army Knife
The architectural philosophy behind keeping Scout, Ticker, Matcher, Quill and Sage beautifully separate.
The New Gold Standard for AI Governance and Responsible AI
The New Gold Standard for AI Governance and Responsible AI
What best-in-class AI governance actually looks like — and the frameworks organizations need to build it.
Ensuring AI Implementations Are Fair, Explainable, Socially Beneficial and Human-Centric
Ensuring AI Implementations Are Fair, Explainable, Socially Beneficial and Human-Centric
The four pillars that separate responsible AI deployment from checkbox compliance.
Responsibility of Responsible AI for Organizations
Responsibility of Responsible AI for Organizations
Why AI responsibility isn't a team — it's a culture. And what organizations must do to build one.
Responsible AI: The Urgency Is Real, But the World Is Not Lost
Responsible AI: The Urgency Is Real, But the World Is Not Lost
A practitioner's take on why we need to act now — without the panic.
Using the Hippocratic Oath to guide Responsible AI
First, Do No Harm: Using the Hippocratic Oath to Guide Responsible AI
Medicine's oldest ethical principle applied to AI governance — by a practitioner who's seen both sides.
The Race to AGI: Who Is Doing What, Where We Stand, and the SSI Wildcard
The Race to AGI: Who Is Doing What, Where We Stand, and the SSI Wildcard
A clear-eyed map of who's building AGI, what strategies they're betting on, and what the SSI wildcard changes.
The Council of AI: When Multiple LLMs Debate, Everyone Wins
The Council of AI: When Multiple LLMs Debate, Everyone Wins
Why asking one AI is like consulting one doctor — and what happens when you bring in the whole board.
The Agentic Web: what happens to the internet when AI does the browsing?
The Agentic Web: What Happens to the Internet When AI Does the Browsing
Where does Google fit? Where do websites fit? Big, provocative questions about the next era of information.
AGI Is Not the Finish Line — Alignment Is
AGI Is Not the Finish Line — Alignment Is
Less about who gets to AGI first. More about what happens the morning after someone does.
AI Models Sometimes Confess Their Own Misbehaviour. Nobody Is Required to Read It.
AI Models Sometimes Confess Their Own Misbehaviour. Nobody Is Required to Read It.
A window into how AI systems think is open right now. It will not stay open forever. Here is what good developers do about it, and what you should be asking your vendors.
20 Years, 100 Clients, One Rule
20 Years, 100 Clients, One Rule
The micromanager, the visionary, the numbers person — and what each one taught me about putting clients first.
More leadership & people stories coming...
More stories from 20 years of people-first leadership
More articles about people, teams, clients, and the human side of working in AI are on the way.