I learnt the product through the work it had to support

The first implementation I worked on had two versions of completion. Our internal plan said the configuration was finished. The customer's operations manager said the system could not yet produce the exception list her team used each morning. Both statements were defensible within their own definitions. Neither helped the people waiting to start work. I was a junior analyst, and my task was to trace the difference back through workshop notes, a sales commitment and a product assumption nobody had written down clearly.

I had joined Sovarel Digital in August 2000 after completing my computing degree at Merrowell Institute. The firm implemented business software for regional manufacturers. I spent my first years documenting workflows, preparing test data and sitting with users who could explain their exceptions more clearly than their standard process. I learnt that a demonstration of the happy path tells an implementation team very little about the work that will occupy its first month after launch. The useful questions concerned incomplete orders, disputed quantities and information arriving in the wrong sequence.

In July 2004 I became a consulting manager at Sovarel. Leading several engagements exposed the commercial side of those questions. Customers wanted a clear scope; sales teams wanted a compelling commitment; delivery colleagues wanted requirements that would remain stable long enough to build. I began reviewing acceptance terms before the proposal was final. This sometimes made an offer less convenient to negotiate. It also reduced the occasions on which a delivery team was asked to resolve a disagreement that had been created before it joined the conversation.

I moved to Arclen Foundry as product director in January 2008. The company built workflow software rather than implementing a broad range of products. I wanted to work closer to the decisions that determined what every customer would receive. We had a strong engineering team and a growing collection of customer-specific branches. Some branches reflected real requirements the main product had not yet understood. Others preserved an urgent promise long after its original sponsor had left. I learnt to distinguish a useful signal from a commitment that should not become permanent architecture.

In September 2013, I joined Nerivane Systems as regional vice president for Southeast Asia. The role combined commercial responsibility with oversight of implementation outcomes. We were selling into organisations with different languages, procurement habits and operating maturity. I resisted the idea that regional expansion could be measured only through contracts signed. A customer needed a workable route to adoption, including its own data preparation and decision-making capacity. We changed the point at which an engagement counted as ready to start and made unresolved customer dependencies visible before allocating the full delivery team.

My appointment as chief operating officer in July 2017 extended that work across the company. I led implementation, customer success and business operations, with 520 colleagues and a US$62 million operating budget at the midpoint of the role. I also helped prepare the business for a change in ownership. The process forced us to reconcile several descriptions of the company: the recurring-revenue record, the contractual obligations, the supported product estate and the work experienced teams were quietly carrying outside their formal plans. Each description was useful; together they showed where our economics depended on unrecorded effort.

I became chief executive in January 2021, after an institutional investment had changed the board's composition. My remit was to grow the enterprise while making its product and service commitments more repeatable. At June 2026, the business recorded US$148 million in annual recurring revenue and served 615 enterprise customers. I distinguish that recurring-revenue measure from recognised revenue and cash collected. It helps describe the subscription base, but it does not explain the cost of implementation or the quality of a customer's eventual use.

One of my earliest decisions was to reduce the number of separate product branches. I knew from experience that a technical consolidation could become a commercial breach if the customer was treated as an obstacle to be migrated. We mapped contractual commitments, tested alternatives and gave account teams time to discuss the changes. Four customers chose to leave at agreed renewal points. That was a consequence of the decision, not a detail to hide inside a product-success story. Nine branches remain, with named owners and explicit reasons for their continued support.

More recently, an automated document assistant has made us revisit the meaning of value in software. A system can produce an answer quickly while leaving a customer with expensive verification work. I asked the team to measure correction and unresolved exceptions alongside output volume. The first release became narrower than the sales plan had assumed. I accepted the deferred bookings because the alternative would have encouraged customers to infer a level of reliability our testing did not establish.

I enjoy the point in an executive discussion where a vague disagreement becomes a specific choice. Engineering may be describing maintainability, sales may be describing a customer's urgency, and finance may be describing the cost of delay. My work is to make those concerns comparable enough for a decision without pretending they are identical. I want colleagues to understand what we have chosen, what remains uncertain and who will return with evidence. That is the contribution I would bring to another board as well as the company I currently lead.