At a glance

CTO and Field CTO for companies of two hundred to a thousand people.

I lead technology as people and organization work. The talent problems and the system problems get solved together, by the same leader.

  • 65 people, $20 millionVP and GM of Data at Xero: engineering, platform, analytics, and enablement.
  • $300,000 to $50,000Yearly cost of Ticketmaster sales reporting after the rebuild, at 30-second freshness.
  • CTO, RadixEngineering and product from September 2025, after an advisory start.
  • $18 millionCompuNet's sale to First Advantage, after I moved its operation onto the web.

XeroTicketmasterLive NationRadix

People and organization

  • Xero. I worked with managers through a change in the performance framework, supported the people it affected, and protected team capacity. Data-organization retention stayed at 95% or higher, and the successor I developed is still in the role. A dozen teams moved onto the shared platform: every model, and every ETL job to dbt or our custom data movement platform. MicroStrategy and the Redshift warehouse were retired, and hundreds of people use the new platform every week. Revenue Operations and 20-25 product analysts became development partners, and both groups stood behind the shared framework within 18 months.
  • Radix. Clearer roles and career expectations, delivery pods, a principal architect I recruited, and one plan that covered staffing, leave, and architecture together. The teams carried on after the engagement.
  • Ticketmaster / Live Nation. I led a data engineering organization of about 45 people.
  • School board. We turned established practice into policy the next board can understand and use.

I have run Scale Data since 2024, building software and the platforms it runs on. That work included an engagement as CTO at Radix in 2025. Before that I was VP and GM of Data at Xero, with a remit of 65 people and a combined-organization budget of about $20 million.

Across those roles, I have worked with people who understand their business but cannot get what they need from the technology around them. Their existing ways of working often contain knowledge that a replacement needs to preserve. I want to understand that with them, and give the people responsible for the result enough room and support to make decisions of their own.

The work

1995 to 1997, Lake Havasu City

LINKnet

I judge a technical shortcut partly by who inherits the inconvenience. If I save money on hardware but leave the person at the desk unable to answer a customer's question without finding me, some of that saving has been paid for by their time and confidence. There are circumstances in which that is the only affordable starting point, but I should recognize the dependency and make a plan to reduce it.

I build companies so they stop depending on their founding engineer. Early on a business may need that person intensely; as it grows, other people's ability to run and improve the service is part of what the founder has to build.

December 1999 to May 2004, Lake Havasu City

CompuNet

Before I back a modernization, I ask what the business will be able to do afterwards. Replacing an unsupported platform or cutting risk can be reason enough, and I say so plainly. A new interface on the same constraints is a different proposition from changing how information enters the business, how customers request it, and which products can be sold.

The CompuNet work matters to me because those changes were connected. The opportunity came from treating the application as part of the information business and understanding how the people on both sides could participate. The commercial timing was part of the technical judgment, particularly in an industry where being years earlier changes what is possible.

August 2004 to January 2021, Scottsdale

Ticketmaster / Live Nation

In a crisis I make the cost to the team visible, and I leave behind a system people can sustain once the pressure passes. The pandemic delivery matters because of the platform and data products the teams delivered under those constraints, including a development framework that made their existing expertise useful quickly.

I would be equally direct about whose design survives. If the people responsible for operating a system can more readily scale and maintain a colleague's approach, the time I spent on mine does not create an obligation to keep it. A technical leader has to make it possible for that comparison to be candid, particularly when the person arguing for the other approach reports to them.

January 2021 to July 2024, Melbourne

Xero

When people don't adopt a platform, I first ask whether the platform understands their work. The people who kept the function running get a clear account of what improves, what gets harder, and how they'll be supported. A central team's architectural preference does not settle those questions.

When a team's capacity is moved, I revisit the date or the scope as a leadership decision, so the contradiction doesn't land on the people least able to change it. When I hand over a program, I want my successor and the teams to inherit the authority and support to make those decisions, as well as the architecture and roadmap.

Advisory, then CTO. July to December 2025.

Radix

An architecture proposal is incomplete until it explains how the organization can get from the existing system to the proposed one while continuing to serve customers. The target diagram may be elegant, but the route depends on business rules, dependencies, and people whose knowledge is often concentrated in the parts everyone most wants to replace. I would want the proposal to show where that knowledge is preserved and how the engineers can verify each change.

A plan that schedules the same people against product delivery and migration as though each has their full attention is making a promise they cannot keep. It is the leadership team's responsibility to resolve that conflict before treating it as a delivery failure. The engineers need room to disagree with the plan without that disagreement being read as a lack of ambition.

2024 to present, Phoenix

Scale Data

The ability to generate more code increases the importance of deciding which changes deserve to exist. If engineers become responsible for reviewing a growing queue of plausible implementations while the people asking for them remain unable to try the result, much of the supposed gain may have moved into someone else's workload. I want to understand that whole process before treating generation speed as a useful measure of development.

I also do not think a successful test run is enough to decide that a repair helped its user. The tests can establish behavior we already knew to check; the person trying to complete the work may reveal a misunderstanding in the request or the proposed fix. That person needs a meaningful place in the workflow, with human responsibility for what reaches production and a way to ask for another attempt.

How I run things

If a team only feels safe telling me what it thinks I want to hear, my confidence in the plan is based on incomplete information. I need to consider my own part in that before asking people to be more candid. The response to an inconvenient finding teaches the team whether it is useful to bring the next one, particularly when it challenges a decision I have already supported.

I also think leaders should be careful about praising exceptional effort when it repeatedly compensates for decisions they have left unresolved. People can choose to do difficult, urgent work and take pride in it; that does not make their willingness an unlimited planning resource. I want managers to be able to discuss capacity, development, and the effect on the people doing the work without having to present those concerns as delivery excuses.

Outside the day job

If the people responsible for an organization's work have to come through me for routine changes, I have made myself part of their operating constraint. I make sure the teams I lead can make routine changes without me: tools they understand, a way to review consequences, and the knowledge they'll need when someone else takes over. That still leaves a role for technical support, without making every decision a technical request.

AI across the company, not only in engineering

People. ArtsEd West's cofounders publish through Google Sheets while an agent stages their changes for review. Stout Tent's crews work from an app built with them. The person who reports a production problem will be able to try the proposed fix and ask for another attempt; that part is being built now.

Systems. Agents work in isolated environments to bounded specifications. Production credentials and promotion stay outside their control, and a human reviews what ships.

The pattern

The word harness comes up often now in conversations about AI agents and harness engineering. I have been thinking about harnesses for intelligence since long before I was directing agents, and for me they begin with clear information: what the work requires, how the platforms and tools support it, and where someone has room to exercise their judgment.

A document defining roles and responsibilities for data, software, machine learning, and QA engineers at different levels is a harness in that sense. It gives people context for what is expected of them and how their responsibility can develop. I also want people to be able to see the expectations for other roles, so they can understand a different kind of work and have a go at it if they want to.

When I spend time helping people understand a platform and how it is meant to be used, I also learn where the explanation is unclear and where the system is unnecessarily difficult. Those questions feed back into the design, making the harness clearer and simpler as it matures.

I am doing the same work with AI agents, choosing among frontier model families and effort levels for the job at hand. I adjust the context and direction as I see how each handles the work, much as I have had to learn when different people need more support, a challenge, or room to try another approach.

WhenWhoWhat they could doWhat supported it
2000Staff, carriers, and brokersCollect data and request reports through a web interfaceShared data, report generation, and automated delivery
2004 to 2021Ticketmaster clientsMore detailed financial, sales, and attendance reportingA rebuilt data model and reporting pipelines
2019Teams managing busy ticket salesUse current sales numbers during an on-saleChange capture, reporting rules, and validation
2020Ticketmaster engineersDevelop data products in SQL notebooksSpark, scheduling, recovery, and deployment
2022Xero analysts and product teamsBuild and own their modelsdbt, the vault, ownership metadata, and shared support
2021 to 2024Xero product teamsUnderstand user actions and background product activityInstrumentation, event ingestion, and customer enrichment
2025Radix engineering and product teamsPlan product changes and a cloud migration togetherArchitectural boundaries, capacity planning, and clear responsibilities
2025Radix engineers and managersUnderstand expectations and discuss career developmentShared roles and responsibilities across disciplines and levels
2026Stout Tent operationsCoordinate crews, travel, and equipmentA replacement application and shared hosting platform
2026ArtsEd West cofoundersPropose and review publishing changes through Google SheetsAn agent that stages their changes for review
2026Coding agents and the engineers directing themWork on bounded changes in parallelIsolated development environments, tests, and review

What I want the work to leave behind

The point of this work is that an analyst can answer a question while it still matters, a crew manager can trust the plan in front of them, and an engineer can change a system without needing its author in the room. I want the people doing that work to have a say in the systems they depend on, and the team building those systems to have the authority, support, and technical understanding to keep making them better.

The architecture, the budget, and the way a team is managed all contribute to that result, so I want responsibility for how they fit together. If that is the kind of work you need help with, tell me what your people are trying to do and what is getting in their way.

dan.franks@gmail.com Work with me LinkedIn