Skip to content
Michael O'Sullivan, founder of Decaf Data.

About

I am Michael O'Sullivan, founder of Decaf Data.

I have spent the last decade running data and AI enablement engagements as an independent consultant, working with organisations across superannuation, infrastructure, financial services, professional services, and consumer brands. Decaf Data is how I do that work today. I lead every engagement personally. The person you meet in the first conversation is the person who delivers the work. Here is what that actually looks like in practice.

The work

What I actually do.

My work splits roughly three ways. There is the early-stage strategy work, where an organisation knows it needs to do something about data or AI but has not yet shaped what. There is the foundations work, where the strategy exists and now needs to become a working operating model, with frameworks, ownership, controls, and playbooks the team can pick up and run. And there is the longer arc, where clients keep me alongside as the enablement matures.

Most engagements start in one of those modes and quietly move into another. A strategy piece becomes a foundations build. A foundations build becomes a multi-year partnership. The shape of the work changes as the organisation's needs change, and I prefer it that way. Enablement is not a project you ship and walk away from.

The day-to-day work itself is mostly conversations. With executives about what they are trying to enable. With domain leads about how their data actually flows. With stewards and owners about what good looks like for their patch. The frameworks and operating models come out of those conversations, not from a template.

Track record

Examples of recent engagements.

A selection of recent work across superannuation, infrastructure, insurance, legal services, and consumer brands. Each engagement shaped how the organisation thinks about and runs its data and AI enablement today.

Superannuation

Working with a major Australian superannuation fund on their data and AI enablement. A multi-stream engagement spanning governance, strategy, cyber security, and operating model design. The work touched most of the data function and the operating model it established is still how the organisation runs today.

Infrastructure

Working with a national infrastructure operator on a multi-year data and AI enablement engagement, structured around bi-annual deep-dive sessions and ongoing socialisation. Maturity has continued to compound year on year rather than stalling after the initial build.

Insurance

Working with a lenders mortgage insurer, stood up the organisation's data and AI enablement capability from a standing start. The work included governance and data quality frameworks paired with a programme of lunch and learn sessions to build internal capability. The frameworks have been carried forward by the in-house team since.

Legal services

Working with a top tier law firm, a focused assessment of the data and AI enablement landscape, delivering the executive a clear, prioritised view of where the practice was strong, where the gaps were, and what to invest in first.

Consumer brands

Working with a major consumer brands group, strategy work shaping the direction for data and AI enablement across the group's portfolio of businesses. The work gave leadership a coherent plan to execute against.

Clients named on request, where I have permission.

Beyond consulting

I also teach this work.

Alongside client work, I have spent years training data professionals in data management best practice. I have trained over 500 practitioners across Australia and the wider region in foundational and advanced data management topics, including as a trainer accredited by the EDM Council for the Data Management Capability Assessment Model (DCAM).

Teaching keeps the practice honest. It is harder to oversimplify a framework when you are explaining it to a room of practitioners who will ask hard questions.

Trainer in data management practices · 500+ practitioners trained

How I work

Three things I have learned.

Governance has to be tied to delivery and outcomes.

I have seen plenty of governance frameworks die in a SharePoint folder because they were built in isolation from the work they were meant to enable. The frameworks I build are shaped around the use cases, decisions, and outcomes the business cares about, not the other way around.

What I leave behind has to be picked up and operated.

If the operating model only works while I am in the room, I have failed. I build playbooks and operating models that read like instructions, sized to be picked up by the team accountable for them. The test is whether they can run it without me.

The senior person sells, scopes, and delivers.

If you meet me in the first conversation, I am the person doing the work. I do not run a model where a partner sells and an analyst delivers. You get me, throughout.

Background

A bit about me.

I am from Ireland originally, and have spent over twenty years in data and information management, working across financial services, government, and professional services. I founded Decaf Data in 2019 to focus specifically on data and AI enablement as an independent practice, and it has been my full-time work ever since.

I am based on the Central Coast of New South Wales and work with clients across Australia, onsite, remotely, or wherever the engagement needs me to be.

Central Coast, New South Wales

Let's talk.

If any of this resonates, I would rather have a conversation than send a proposal.

Book a conversation