Maker of AI apps & tools · Sydney, Australia

I design and build AI apps, end to end.

Over the last two years I have built 25+ working apps and tools for private knowledge search, voice, 3D, live events, and everyday work. A few of them are below.

25+
apps & tools built
~2 yrs
building, hands-on
Web · Desktop · Mobile
across platforms
About

I'm a maker. I see a slow task, and I build a tool to fix it.

I work mostly on AI-powered software. Three things matter to me in everything I build:

Private by default

Many of my apps run fully offline on local AI, so data never has to leave your own machine.

Simple to use

I keep tools clean and clear, so a person can pick them up and get value without a manual.

End to end

I take an idea all the way through design, build, and ship, on web, desktop, and mobile.

Selected projects

A few of the bigger projects I've built.

Every screenshot below is the real app running.

01 · Flagship

NodeBook

A private "chat with your documents" platform for a whole company, like NotebookLM, but it runs 100% offline on the local network, so no data ever leaves the building. People ask questions in plain language and get answers with the exact sources.

Hours of manual searching → answers in seconds, with sources, and 100% offline.
  • Answers with citations: every reply links back to the exact document and page.
  • Many ways to ask: Search, Summarise, Compare, Extract, and Define modes.
  • Your files: drag-and-drop PDF, Word, and text; answers stream back live.
  • Access layers & tags: role-based control, so each person only sees what they're allowed to.
  • Admin dashboard: users, documents, AI-model choice, and live search analytics.
  • Meeting assistant, team messaging & a built-in tutorial so new people learn by doing.
RAGLocal LLMs (Ollama)Vector searchFastAPINext.jsTauri
nodebook · local network
NodeBook answering a question with sources NodeBook admin dashboard
02 · Product

3DP Ocean

A 3D-model marketplace where AI does the whole job for you. You just give it one thing (a single photo, a few photos, or even a few words) and it builds a full-colour, print-ready 3D model. No 3D skills needed.

A multi-step 3D workflow → one input. A photo or one sentence becomes a print-ready model.
  • One simple input: a photo, a few angles, or a text description with a tiny AI Q&A to refine the idea.
  • It makes the angles for you: in the background, AI cleans your photo into a studio shot and generates the missing front, side, back, and 3/4 views by itself.
  • Smart auto-router: it works out the subject (object, animal, figure) and picks the best path automatically.
  • Its own 3D Studio: edit the model right in the browser (no Blender) and remove bad parts, snap on better ones, then merge everything into one clean printable mesh.
  • Print-ready: full colour, quality tiers, model scales for miniatures, and one-click part separation.
  • A real product: accounts, credits, membership tiers, private models, and a voting-based marketplace.
Next.js 15React Three FiberIn-browser 3D editorImage / text → 3DStripePostgres
3dpocean.app
3DP Ocean
03 · Web + Mobile

TempoWhisper

A full web and mobile app for running community events, built around your voice. Tap to speak, and it creates the event for you. Live online at tempowhisper.com.

Create an event by voice in seconds, in 16 languages, on web and mobile.
  • Voice-first scheduling: say it out loud and the event is made, with the date, time, and details filled in.
  • Community calendar: browse, submit, and manage events; track your own submissions.
  • Built-in image generation: make a poster image for any event in seconds.
  • Google Calendar sync, email invites and reminders, and event approvals.
  • Accounts & billing: Google / Microsoft sign-in, Stripe payments, and team workspaces.
  • 16 languages: fully translated, on web and mobile.
LaravelReactVoice → eventStripeGoogle Calendari18n (16 languages)
tempowhisper.com · dashboard
TempoWhisper voice dashboard with upcoming events
04 · Desktop

Seminar Timeline

A desktop app that runs a whole live, multi-talk event from one timeline, across two screens, the kind of tool AV teams pay a lot for. The operator controls everything; the audience only ever sees a clean show.

One operator runs a whole multi-talk event, replacing paper forms, email, and pricey AV tools.
  • Two screens, one operator: the controls on the laptop, the slideshow on the audience monitor.
  • Drag-and-drop timeline: build the day as slots (talk, break, intro, ad) with a PowerPoint and audio in each.
  • One-click playback: launches PowerPoint fullscreen, starts the audio, and gives Play / Pause / Next / Back / End.
  • Live preview of exactly what the audience sees, with delay tracking so the event stays on time.
  • Breaks done right: break images, background music, and countdowns between talks.
  • Never a black box: every file is saved on disk, so nothing is ever lost, even if the app closes.
ElectronReactTypeScriptDual-monitorZustand
seminar timeline · operator console
Seminar Timeline operator console
And many more

Smaller tools I build to make daily work faster.

When something is slow or boring, I build something to fix it. A few examples:

HoverVox

A floating voice-to-text bubble: press it, speak, and your words are typed into any app.

LocalLens

An AI agent that tests web apps for me, clicking through the whole app and reporting problems.

3D & design integrations

I connected Blender, Revit, and ComfyUI to AI, so they can be driven automatically.

InSite

A newsletter website you edit from inside itself. Admins and members sign in, and admins change any page live in the browser, with auto-backups.

MailCraft

A desktop app to build polished email templates: pick a style, arrange the blocks, add images, and export ready-to-send HTML. Sign-in saves each person's work.

Personal AI assistant

A personal assistant that helps with everyday tasks, plus 15+ more small tools.

Get in touch.

Want to know more about any project, or see a live demo? I'm always happy to show my work.