AI systems builder · Queensland, Australia

I turn messy, high-stakes workflows into AI systems people can trust.

I help founders and businesses turn difficult workflows and product ideas into useful, production-minded software—from strategy and prototyping to agent systems, integrations, controls and deployment.

Discuss your workflow View product portfolio

0110+ years solving operational problems

02Products across six different markets

03Strategy through to production

Where the leverage hides

Your workflow is already telling you what to build.

The strongest AI opportunities often look unglamorous at first: a document copied twice, a decision nobody can trace, or a queue that depends on one person remembering what happens next.

01

Knowledge is scattered

Critical context lives across documents, inboxes and experienced team members.

02

Work is repeated

People re-key, reformat and re-check information instead of moving the decision forward.

03

Trust is fragile

A clever answer is not enough when the team needs sources, permissions and a clear approval path.

04

The opportunity

Turn the workflow into a focused system that helps people make better, faster and more visible decisions.

Flagship system · Financial advice

Sombra

Compliance-aware AI for financial advice, shaped around reviewable work rather than black-box output.

The product principle

Source visibility and practitioner judgement belong inside the workflow.

Sombra is an evolving system for turning advice material into structured work. The boundary below is deliberate: it distinguishes what is working in production from what has been designed or explored.

In production

File-note and document workflows

Working software supports structured file-note capture and document workflows with source-aware review and human sign-off. These are the production workflows today.

  • Structured information capture
  • Document and source context
  • Review before completion
Designed and explored

Broader agents and controls

Concepts and prototypes explore specialist-agent orchestration, policy controls, audit visibility and escalation paths. These are not presented as deployed production capability.

  • Specialist agent roles
  • Policy and permission layers
  • Escalation and audit concepts
Read the full Sombra case study

A practical service ladder

From ambiguous workflow to useful system.

Start at the point your team needs. Each stage creates something concrete enough to test, review and build on.

  1. 01

    Workflow diagnosis

    Map the work, decision points, source material, risks and highest-value place to begin.

    Find the leverage
  2. 02

    Product strategy + prototype

    Turn the opportunity into a clear product shape and a prototype people can respond to.

    Make it tangible
  3. 03

    Controls + integration

    Design permissions, sources, human review and connections to the systems already in use.

    Build trust in
  4. 04

    Delivery + iteration

    Ship the narrow useful slice, observe the real workflow and improve from evidence.

    Move into practice

Beyond the flagship

The same product discipline, in very different markets.

Sombra proves the regulated work. These five products show the range—from consumer apps and education media to private infrastructure and internal production systems.

02Built platform · active validation

Tend

An owner-controlled pet passport connected to booking and facility operations.

03Working product platform

Stories Alive

A multimodal teacher platform for coordinated stories, songs, video and lessons.

04Personal MVP · review pending

Agent Layer

A personal health-data MVP for encrypted, user-controlled agent context.

05Advanced pilot · active validation

Expat Sidekick

A document-to-action assistant built around visible evidence and editable facts.

06Live internal product

Website Factory

A deployed workflow for site audits, redesigned previews and campaign operations.

See the problems, architecture and product decisions behind all six systems.

Explore the full portfolio

Motion as explanation

Complex systems become clearer when you can see the logic move.

Short motion studies can make product behaviour, safeguards and decision flow understandable before a system is built.

Loop / 07s
01

Trust Loop

Source → system → human decision

Study / 07s
02

Signal Brief

Turning scattered inputs into a reviewable brief

Workflow-first method

Build around the decision, not the demo.

A useful system fits the people, information and controls already surrounding the work.

01

Observe the real work

Start with artefacts, hand-offs, exceptions and the people closest to the workflow.

02

Define the decision boundary

Be explicit about what the system can suggest, what it must show and what a person decides.

03

Make trust inspectable

Expose sources, uncertainty, permissions and approval steps in the product experience.

04

Ship a narrow useful slice

Put one valuable workflow into practice, learn from real use and expand deliberately.

About Ryan

A product-minded builder who understands the operational side.

I have spent more than a decade solving operational problems and turning ambiguous work into practical systems. Today I combine product strategy, AI prototyping, workflow design and hands-on delivery—especially where trust and human judgement matter.

I work best with teams who know there is a better way to run an important workflow, but need help shaping the opportunity and making it real.

Useful questions

Before we talk.

What kinds of workflows are suitable for AI?

The strongest candidates usually involve repeated document work, fragmented internal knowledge, manual handoffs or decisions that benefit from consistent preparation. The goal is not to automate everything; it is to remove friction while preserving the judgement that matters.

How do you decide what should and should not be automated?

I map the current workflow, the information it depends on and the consequence of a wrong action. Low-risk, reversible work can move quickly. Sensitive decisions, external actions and uncertain outputs get explicit review, approval or escalation points.

Can you work with sensitive or regulated information?

Yes. My work has included regulated financial-advice workflows, multi-tenant isolation, Australian data-residency requirements, approval controls and traceable system activity. The exact architecture depends on your risk, policy and data constraints.

Can you connect AI to our documents and existing systems?

Yes. I build around the systems and information your team already uses, including company documents, templates, customer records, APIs, webhooks and operational tools. The useful outcome is a connected workflow, not another isolated chat window.

What does a prototype sprint deliver?

A focused sprint turns one meaningful workflow into a working, testable version. Depending on the problem, that may be a document review tool, internal assistant, structured drafting workflow, dashboard or agent-led research process.

Do you train machine-learning models?

My focus is applied AI rather than foundation-model research. I combine strong product thinking, frontier models, retrieval, software, integrations and controls to make AI useful inside a real business process.

Have a workflow in mind?

Let’s find the useful system hiding inside it.

Tell me what your team is doing today, where the friction lives and what a better outcome would make possible.

Discuss your workflow on LinkedIn