Practical writing on using AI well

What worked, what failed, and what helped me move from pilot to production.

New posts ship on the blog; the weekly digest catches what you missed — one email, practical only.

Nathan Petralia at HKU

Nathan Petralia

I have spent two decades leading digital and commercial programs across APAC. Today I build AI products and use agents daily across consulting, practice building, go-to-market, delivery governance, and operational leadership.

About me →

Start here

New to the site? Pick one path — each opens a guide or series hub.

Browse all writing →

Reading paths

Multi-part guides — start at the hub, then follow the series in order.

Latest

New this weekSteel filing vault with one drawer open showing glowing data cards instead of software logos.
Hybrid

The CDP Your Agents Need Is a Folder Contract

Best forMarketing and program leads running agent pilots who need governed customer context without buying another CDP seat

Composable martech plus a folder contract: governed customer-context paths for agents—when CDP, lake, or hybrid stacks fit, and best practice for commerce and marketing.

Continue reading →
New this weekRetail floor at dusk with wireframe garment on mannequin and code-like lattice reflected in storefront glass.
Strategic

Shopify Hydrogen vs Online Store 2.0 When AI Is the Front End

Best forEcommerce operators and tech leads choosing Shopify storefront architecture when AI assists design, copy, and component generation

When AI generates storefront UI and copy, the stack choice shifts. Hydrogen offers component control; Online Store 2.0 offers speed. Here is how I decide.

Continue reading →
New this weekCinematic 16:9 of three generic chat kiosk screens inside a Hong Kong tram at dusk, disconnected speech bubbles, no logos.
Strategic

Hong Kong Customer AI Is Still Mostly a Label

Best forCX leaders, digital banking heads, and public-sector service owners benchmarking Hong Kong front-line AI against policy rhetoric

Three Hong Kong touchpoints in one afternoon—IRD, Standard Chartered, HSBC—failed the same simple tests. Policy and press releases run ahead of what citizens and customers actually experience.

Continue reading →
New this weekIndustrial conveyor belts move scheduled packets while a lit desk handles one open folder for judgment work.
Hybrid

n8n vs My Cursor and Obsidian System: Scheduled Automation vs Judgment Work

Best forOperators comparing n8n to a Cursor plus Obsidian harness who need a clear line between scheduled integrations and file-grounded agent judgment

I retired n8n last year for AI-coded microservices on petralian.com. Cursor plus Obsidian still owns judgment work. Here is when each layer wins.

Continue reading →
Smartphone with chat glow on an armrest and a laptop displaying a folder tree on a table.
Hybrid

OpenClaw vs Cursor in 2026: Ambient Messaging Agent vs File-Grounded IDE

Best forAnyone who read OpenClaw hype and uses Cursor who wants a clear comparison between ambient messaging agents and file-grounded IDE work

OpenClaw is my mental model for ambient agents on WhatsApp and Telegram. Cursor is where files, rules, and shipping live. Here is the 2026 split.

Continue reading →
Phone on a kitchen counter with message bubbles and a desk with open document folders in the background.
Hybrid

Hermes vs Cursor in My Setup: Life Agent on the Server, File Work at the Desk

Best forAnyone comparing a hosted messaging agent with Cursor who wants a clear split between ambient life admin and file-heavy project work

Hermes is my hosted life agent on Telegram and family WebUI. Cursor is my file-grounded desk agent. Here is when each wins in 2026.

Continue reading →