Skip to content
Grant Show

AI practice

An AI practice 20 years in the making

I don’t think anyone is truly AI-native. But AI has become the medium I work in, the way I design, write, research, build, and lead.

Throughout my career I’ve moved from designer, to full-stack engineer, to serial entrepreneur. In that sense I’ve been getting ready for AI my entire working life, and by that measure, I’m about as AI-native as one can be.

The public Outlook website: a scattered photo collage behind the line 'Flagship conferences for product, design, and leadership'.
Website — the public home of The Outlook

How I work with AI

I use AI across all of my work, not just to write code. Here is where it shows up most.

Design and prototyping

Moving from idea to working interface in hours, and testing in the real material instead of a static mockup.

Writing and narrative

Drafting, structuring, and sharpening strategy, copy, and stories. This site included.

Coding and shipping

Building and deploying production apps across APIs, data, and automation, end to end.

Research and analysis

Reading large, messy systems quickly, finding the pattern, and turning it into a decision.

Critique and review

A second pair of eyes on craft, security, and edge cases before anything reaches a user.

Systems and strategy

Modelling complex domains and operating models so the work stays coherent as it scales.

Teaching and speaking

I also teach and speak about AI

Building the practice is one thing. Helping other people build it is another. I run masterclasses on AI in design, and I speak on where it leaves human judgment most valuable.

AI for Designers: Foundations

Masterclass

AI for Designers: Foundations

The Outlook · Masterclasses 2026

A full-day, hands-on workshop on how AI fits into modern design work, and how to use it confidently across the whole end-to-end process.

Beyond AI: Human insight as the advantage

Keynote

Beyond AI: Human insight as the advantage

Digital FSI 2025

A keynote on what stays valuable as AI spreads through an industry: human insight, judgment, and the trust that sits underneath good decisions.

Case study

A seven-app product ecosystem, built with AI

Over the past year I designed and shipped seven production apps, on one shared design system, that now run a real events and media business.

Seven production apps One shared design system Designed and built solo
Session and program management for a multi-track conference, scheduled by drag and drop.
Content Session and program management for a multi-track conference, scheduled by drag and drop.
Live event financials: budgets, sponsorship, costs, and net position against target.
Events Live event financials: budgets, sponsorship, costs, and net position against target.
Event planning and management, with automated seating across tables and rounds.
Attend Event planning and management, with automated seating across tables and rounds.
Sales charting and comparison across events: revenue, tickets, and pacing to date.
Events Sales charting and comparison across events: revenue, tickets, and pacing to date.

CRM

The people layer: contacts, sponsors, and partners across the business.

Events

Budgets, P&L, and automated billing for every event and season.

Attend

The attendee app: passes, schedules, and the on-the-day experience.

Content

Presentations, schedules, and self-serve for speakers and facilitators.

Hub

The integration backbone that keeps every app in sync.

DAM

A digital asset manager, with AI-assisted tagging and search.

Website

The public marketing site, on the same shared design system.

COPE

Create once, publish everywhere: each app is the canonical home for one kind of data, projected to the others through the hub. One source of truth per domain keeps the model drastically simpler, and always ready for AI.

Under the hood

A real, modern stack

Not low-code glue. The same tools a product team would reach for, designed and maintained with AI in the loop.

Interface
React, Tailwind, and shadcn/ui, on one shared design system used by every app.
Data and content
PocketBase and Payload CMS, with Go services for the heavier domains.
Integration
An event-driven hub (outbox and projections) that keeps every app in sync.
Infrastructure
Postgres, SQLite, and object storage, deployed on Fly.io.
Mobile
A React Native app, built with Expo, on the same design language.

Integrations

Wired in, not just connected

These run deep. Money, identity, ticketing, and contact data move two ways through the event-driven hub, and the agents read and write across them: applying transactions in Xero, sending email as me through Graph, and reconciling Stripe payments as they land.

  • Stripe logo

    Stripe

    Payments and webhooks, with automatic recovery

  • Xero logo

    Xero

    Two-way sync, with agent-applied transactions

  • Humanitix logo

    Humanitix

    Live ticketing and attendee sync

  • Microsoft logo

    Microsoft

    SSO, and email sent as me via Graph

  • Slack logo

    Slack

    Operational alerts and agent control

  • Apollo logo

    Apollo

    Automated contact enrichment

Agents

Building AI into the product, not just with it

The next step past building with AI is building AI in. Across the ecosystem I have made agents that do real operational work, each one scoped, supervised, and earning its place.

A Slack thread where the Penelope agent generates and confirms content fields for a record, derived from existing material.
Slack An agent writes and confirms a record's content fields, derived from what already exists, and reports back exactly what it did.

An event producer

Reads every transaction, attributes it to the right campaign with high confidence, and applies it directly in the accounting system. Routine financial work, done before I look.

A digital twin

Drafts and sends email on my behalf, as me, inside a window where I can still step in. Introductions and outreach that would otherwise never get sent.

A support and operations crew

A small roster of named agents that handle support, reconciliation, and operations across the ecosystem, each scoped tightly to one job.

What I bring with AI

I'm a leader with real technical depth, fluent in AI as a working medium, operating at the level where design, product, and the business meet.

What I demonstrate here is exactly that: the technical depth to design and ship production systems myself, AI fluency used across building, research, critique, and strategy, and the judgement to operate where design, product, and the business meet.