Blog
Generative AI: A 'digital copilot' for treasury
4 Minutes
4 Minutes
Generative AI offers exciting possibilities
Since generative artificial intelligence (gen AI) capabilities entered the mainstream in 2022, they’ve made headlines almost daily. Many professionals are experimenting with platforms like ChatGPT and Bard for everything from researching competitors to crafting reports to finding the funniest dad jokes.
Beyond novelty, however, is there a place for generative AI in corporate treasury? Yogaraj (Yogs) Jayaprakasam, Chief Technology Officer at Deluxe, believes there is. In a podcast with Strategic Treasurer, he shared three ways for finance professionals to put these cutting-edge capabilities to work right now.
AI's potential to change the trajectory of business
Generative AI represents a new inflection point for business, according to Jayaprakasam. That’s because its capabilities far exceed those available just a few years ago.
While traditional AI excels at analyzing massive amounts of historical data, generative AI adds the ability to create new content. It starts to reason and “think” much like humans do. This unleashes far more potential. This shift empowers individuals with no knowledge of computing or coding to interface with AI tools because they communicate with human language.
“It means you can interact with the machine without having a mouse, website, mobile phone, or screen as the interface,” Jayaprakasam explained. “It also means the machine itself can start to guide you through the process.”
Arming treasury with 'a really smart intern'
He sees several strategic advantages for organizations that harness generative AI, including:
- Reimagining products and services
- Accelerating reporting
- Improving productivity
- Freeing staff for higher value activities
Jayaprakasam describes adding generative AI to the team as “hiring a really smart intern.” Given the right prompts, it can gather information quickly across vast amounts of data. It can bring back a lot of new ideas. However, just like an intern, companies still need a subject matter expert (SME) to review, validate, and refine the final output.
Three ways finance professionals can leverage new AI capabilities right now
While generative AI continues to evolve, Jayaprakasam encourages treasury to start using it today.
“Treasury has always been thinly staffed; there are many tasks that organizations wish they could get to,” he explained. “Generative AI represents an area where there’s interest and opportunity for everything from increasing automation, to managing AR and AP more powerfully, to increasing visibility to counterparty risk and cash flow forecasting.”
Jayaprakasam recommends finance explore generative AI with these three activities:
1. Accelerate and improve training
“Training is the easiest place to start,” Jayaprakasam noted. “It’s a low-risk area where you’re helping someone learn something faster or do their job faster.”
Consider how generative AI can support new employees, entry level staff, or financial professionals moving into new roles as they come up to speed on company processes, historical data, and their overall responsibilities.
Studies show it can be especially helpful in closing proficiency gaps between an organization’s highest and lowest performers. In an AR function, for example, generative AI could support less experienced AR analysts with education and recommendations on topics such as “What is the order-to-cash process?” or “What AR levers can we use to improve our order-to-cash cycle?”.
2. Co-create with an intelligent digital copilot
The next step is pairing the technology with an experienced professional in what Jayaprakasam defines as a “digital copilot” role.
“You still have the main individual doing the job, but then your technology starts to summarize information or automate parts of the process, freeing staff to use their subject matter expertise to generate more value,” he said.
For example, a digital copilot could help treasury increase accuracy in cash flow forecasting.
“Let’s say you asked, ‘What are the associated variables that can drive our forecast up or down?’” he said. “There’s historical information; there might be dependent and independent variables. With generative AI, you can examine far more information, far more quickly, and find more correlations or causations. AI helps get those answers really rapidly.”
3. Strengthen results with proprietary data
A third opportunity for finance is helping generative AI “learn” from company-specific data. AI’s default knowledge base is publicly available data. Businesses that supplement their AI models with proprietary information can strengthen outputs by “teaching” the system with more specific and customized information. The longer the AI models have access to custom data, the better they are able to use it in their analysis and results.
Start exploring to understand AI's pitfalls and opportunities
Not every financial task will be suited to AI, Jayaprakasam cautions, particularly with early versions of generative AI. How well users phrase their questions will also significantly influence the quality of generative AI’s results. He recommends a phased approach.
“Go from low risk to high risk, which is training people, summarizing their work, then getting slowly to assisting with your decision-making,” he said.
Now is definitely the time to get on board. He sees a bright future for treasury and AI together.
“Eventually you can you have the technology help you close the books faster, calculate liquidity positions faster, do capital calculations faster, all of those things,” Jayaprakasam explained. “We need to get involved, start to understand both the opportunity as well as pitfalls, and contribute to it.”
Deluxe Receivables360+
Improve your cash application and forecasting with receivables automation.
RECOMMENDED RESOURCES
