How AI accelerates complex finance work at DRA
While automation handles routine tasks, the true potential of artificial intelligence lies in acceleration, transforming complex financial work that once required weeks into processes completed in minutes, without sacrificing accuracy or oversight.
At Disaster Relief Australia, CFO Mark Welton has moved beyond basic AI applications to tackle sophisticated finance challenges. His second wave of AI implementation focuses on accounting standards and financial modelling, where the technology's impact becomes exponentially more valuable.
"The real power in AI isn't just in generating SQL or writing emails, it's helping CFOs solve actual finance problems faster," Mark explains.
Revolutionising lease accounting
IFRS 16 lease accounting represents one of Mark's most impressive AI implementations. DRA manages complex lease arrangements, including peppercorn leases requiring detailed fair-value disclosures, a challenging area for any finance team.
Mark's initial approach followed traditional methods: building lease models manually using Python and Excel through weeks of painstaking iteration. Today, that same complexity is handled seamlessly through AI automation.
"The GPT gives me the amortisation schedule, disclosure wording, even the journal entries. Then I take the Python code and plug it into Excel or Jupyter. It's done," he describes.
The transformation extends beyond speed. The AI-generated models are repeatable, scalable, and maintain the auditability that finance teams require. However, Mark emphasises that domain expertise remains crucial.
"I tested the outputs against my own models. It's accurate. That's the key. You still need domain knowledge to know if what AI gives you is right."
Instant access to accounting standards
DRA's finance team no longer depends on memory or lengthy PDF documents to navigate accounting standards. Mark's custom AASB GPT serves as an always-available technical resource, instantly answering complex questions such as:
- How should volunteer services be treated?
- What's the correct recognition approach for peppercorn leases?
- How does a new board policy align with existing accounting frameworks?
"It's like having a technical accountant on call. But one that answers instantly, doesn't charge, and doesn't go on leave," Mark notes.
Beyond immediate problem-solving, the GPT functions as a powerful development tool for junior staff. Mark encourages newer team members to query the AI before escalating questions, building their critical thinking skills and professional independence.
Democratising financial modelling
Mark's team now creates sophisticated financial models using a combination of ChatGPT, Google Gemini, and saved notebooks. This approach transforms modelling from a specialised skill into an accessible capability across the team.
He regularly challenges team members with practical exercises:
"I'll say: Here's the scenario. Build me a model. They do it with AI. They learn. They iterate. And they get better every time."
The workflow proves particularly valuable for forecasting, scenario planning, and data reconciliation tasks.
"I use AI to compare tables, match inconsistencies, and generate M-code for Power Query. The output's clean and saves me from manual reconciliation," Mark explains.
Completed Python scripts are stored in a shared Jupyter environment, creating a growing library of reusable finance tools. Team members can adapt existing models by adjusting inputs, ensuring institutional knowledge is preserved and accessible.
Quantifying the impact
The financial implications of Mark's AI implementation are striking. When DRA needed to transition from prescriptive to principle-based policies, external consultants quoted $130,000 for a six-month project. Mark completed the same work using AI assistance in just two weeks.
"We wanted to move from prescriptive to principle-based policies. AI helped us explore that thinking and saved us serious money in the process," he reflects.
The cost savings didn't simply improve budget performance, they enabled resource reallocation to mission-critical operations, demonstrating how effective finance management directly supports organisational goals.
A grounded approach to AI implementation
For finance teams navigating AI adoption, Mark offers practical guidance rooted in real-world experience:
"Treat AI like a junior. Give it clear instructions. Test its work. Then scale what sticks."
The transformation continues
Across accounting standards, financial modelling, and operational insights, DRA's finance function now operates with unprecedented efficiency and capability. From lease accounting to donor analysis, complex work that once consumed weeks now requires minutes.
The technology isn't replacing Mark's team, it's amplifying their expertise and freeing their time for higher-value contributions. In an organisation where every resource must maximise impact, AI has become the force multiplier that makes ambitious goals achievable.
If you're experimenting with AI in your own finance function or want to learn from others who are, jump into The Stack Exchange, our Slack community for finance leaders pushing boundaries.
.png?upscale=true&width=1196&upscale=true&name=Untitled%20design%20(42).png)
.png?upscale=true&width=1036&upscale=true&name=overview%20(4).png)
