The chief audit executive product manager
Water cooler talk at the IIA International Conference 2025
I recently had the pleasure to attend the IIA International Conference in Toronto. Tons of forward thinking professionals catapulting their profession forward. There was a lot of talk about Artificial Intelligence and specifically how teams can start on the path to achieving AI supremacy through experimentation. The reality is many teams are experiencing a dearth of technical solutions for their day to day lives. As a result, forward looking leaders are turning to their internal resources to build these tools. So, if you are an Audit leader and are considering building internal tools (aka someone writing code) for your team, I have a single warning to any Chief Audit Executive or Audit leader.
Building technology is hard
Ok straight to the point - writing code and building technology products is really hard. It requires hundreds, if not thousands of layered good decisions and it costs a lot of money. AI coding assistants haven't made it any easier. Certainly AI now gives developers to do a lot more at a greater speed with better quality. But it doesn't make it easier. Here is a list of things that are still hard.
Security - in the age of Large Language Models, most certainly data will leave your network. The incredible intelligence AI offers to our workflows make them indispensable. So, yes your data will travel across encrypted channels and, yes, most Enterprise LLM agreements will not use your data to train their models. But this is a necessary evil.
Technical debt - Technology starts incurring technical debt on day 1. What is technical debt? Its that Tupperware container in the back of the fridge. Writing features in modern programming languages will become legacy in no time. Packages you depend on will be updated, the core language features will continue to mature and improve, infrastructure will continue to evolve. Its a large platter to keep up with. So, the code your team introduces will be obsolete in no time. Look away and it might stop working.
Product thinking - To build useful technology for humans, we still need to embed ourselves in our user's lives and understand how to translate what they do everyday with technology that makes their lives easier. And this is a skill built up over many years.
Scoping and Prioritization - Deciding what to build — and what not to — is still one of the hardest and most political parts of tech.
Adoption - Its never "just the code". Adoption is a whole other project. And ensuring your team members actually adopt the technology you build is certainly part of the equation.
Team continuity - Lets face it. Your original developers will leave. Devs rotate off, funding dries up, knowledge isn't documented. You may be starting from scratch every quarter.
Maintaining momentum - Software entropy is real. Bugs pile up, roadmaps shift, features get rush, devs lose interest. Keeping a product alive and relevant takes sustained product leadership, not just coding sprints.
The warning
As you can see, technology is still hard. As a technology professional, what would be my big, flashy warning to Audit teams looking to embed building technology prowess in your team? Be wary of cost. Now, Auditors must be technically fluent using AI tools. This has more than just a modicum of truth. As noted many times during the conference, Auditors won't be replaced; just those teams who aren't technology forward will struggle. So grab hold of this bull and learn as much as you can. A few scratchpad notes to keep for your journey.
Building full end to end technology products have bad unit economics. A data science colleague and possibly another developer can have full realized costs of over 200-300 thousand dollars.
Requirements will keep coming. Remember that technical debt we talked about? It will keep marinating like warm mayonnaise in a hot car. Your developers will have to find ways to integrate these features in time constrained environments to meet deadlines. Shortcuts will ensue.
The wrong product features will rise to the surface; a misfire. Given enough time, it will certainly happen. Many data products I saw were still in experimentation phase and its a fragile time. Building the wrong thing can waste months of expensive developer time.
What should Auditors do that are building integrated tech teams?
Data centric
If you need data, learn to get it. Executives need to be trained in dashboards and basic self-service analytics; not just passively reading reports, but exploring the data themselves to ask better questions. And make sure any data analyst folks aren't playing data janitor but rather using their technical skills to do technical work.
Now I get it, data is messy. But that's not an excuse to trap skilled analysts in endless cycles of cleaning spreadsheets, renaming columns, or reformatting CSVs. Build pipelines, use tools like dbt or Python scripts, automate repetitive tasks, and put structure around your data sources.
The goal isn't just to "have data." It's to build a culture where getting answers is fast, reliable, and repeatable. That means investing in the boring stuff like clear definitions, trusted sources, version control, and documentation.
You do need to drive a culture that treats data like a product, not an afterthought.
Limit scope.
When you have the data, you will be tempted to build products around the data. And your job will be to limit scope. What does that actually mean? Build the smallest, simplest solution to solve your problem. This is where you really need to take the reigns and provide leadership and try to find your best Product Manager hat. Left alone, your technology teams will do what they do best - and that is build software with all the aforementioned problems that will creep in.
So your job is to ruthlessly focus:
What exact pain point are we solving?
Who is the user?
What's the simplest version that gets us there both on the UI and backend.
What can we cut, defer, or solve with a spreadsheet or manual workaround?
Remember: the goal isn't to build an app — it's to make your audit process faster, smarter, or easier. The software is just a means to an end.
And if you really want to lead like a product thinker? Be ready to say:
“Not yet.”
“We don't need that right now.”
“Let's measure this first.”
“What if we just use Google Sheets?”
Limit the scope, ship something simple, learn from it and then grow from there.