Top 25 firm Armanino plans to launch a data warehouse service for smaller organizations later this year and has been using AI to develop some of its rollout strategies.
Carmel Wynkoop — Armanino's partner in charge of AI, automation and analytics — explained that big organizations can store and manage large amounts of data in order to analyze trends, optimize processes, fuel AI solutions and much more. However, building and maintaining the infrastructure for this can be very expensive as well as technically complex, so smaller organizations typically lack these capacities.
The idea behind the new service is to level the playing field somewhat by providing these smaller organizations the resources and expertise necessary for a data warehouse of their own. The client would either send Armanino their financial data or integrate it with the firm's directly; from there, professionals would gather the information into a centralized data warehouse—hosted on Armanino's infrastructure—that could be interacted with via a built-in chatbot.
"We are creating a data warehouse as a service… [We are] providing them the reporting and KPIs that larger, more sophisticated organizations—that have a larger IT department and larger revenue—can do but the smaller ones still sort of struggle with a little bit," she said in an interview. "So it's a way for them to get that level of reporting and AI without building their own and having to hire a bunch of people."
The motivation to launch this service line came from observing smaller clients who struggled with getting data-driven insights into their own organizations, which means they tend to ask a lot of questions. Many of them involve the kind of modeling that larger organizations with data warehouse infrastructure can do themselves, such as "if I change this, what will happen?" Wynkoop wondered whether she could empower clients to do this themselves, which then led to the idea for managed data warehouse services.
"We will host it and manage it. Not that other organizations won't do that… but we're basically saying we're going after a smaller footprint of clients here. What we're trying to do is enable firms that want to grow and see their data as a differentiator and competitive disruptor. How do they use that data to make themselves more competitive? Because they would not be able to do that today unless they built their own AI model or loaded their data into an existing public model," said Wynkoop.
She said the firm has people who already know how to build data warehouses, typically for ERP implementations. These would likely be the professionals who would be responsible for the new service line. Asked about the timeline she is working with, Wynkoop anticipated a late Q3, early Q4 rollout.
AI as business developer
One notable thing about this new service line is the role AI played in developing it. AI did not come up with the idea and did not lay out the finer points of how the service would work. Further, AI was not used to develop the data warehouses themselves. Wynkoop said it was mostly used for researching things like the market landscape as well as general brainstorming, making extensive use of the technology in both cases.
"I used Google Deep Research… Then I used ChatGPT to make a marketing plan and had Copilot build me a PowerPoint presentation, then I put that [presentation] back into ChatGPT and Claude. Once I sort of had the business plan and the go-to-market strategy and the PowerPoint, I loaded them [into the models] and asked the three of them to poke holes in it and compare it to other service offerings," she said.
She said that, without AI, this process likely would have taken months. Using AI reduced it to about two weeks.
The go-to-market plan is where the AI fingerprints would be most apparent. Its proposed strategy covered types of clients and industries that would need these services, market share calculations and revenue projections, and possible drivers of interest such as partnering with Microsoft and promoting case studies.
"The biggest piece with use of AI in this scenario was the research, [particularly on the] competitive landscape and in thinking about what features we have to have and what other organizations have that show up in a similar data warehouse offering," she said.
Beyond AI, developing the new service line also required collaboration with the other humans in the firm, particularly IT and legal. No one objected to Wynkoop using AI in her process in and of itself. She said leaders were more concerned about how to make the service safe and secure.
"We're all on board with using AI for research and development of new ideas. I think, because we are going to be hosting client data, one of the things we had to consider and ensure is that we're doing it in a way that is smart and safe, so that is an area we spent a lot of time and focus on and had several discussions with our IT and legal and the whole gamut to make sure all the safeguards are in place for client data," she said.
The service is currently in beta testing. Wynkoop said Armanino is currently searching for clients to help vet the service and make sure the strategy the firm has is viable and valuable.
"II think I think it could be a real game changer for businesses that don't have sophisticated reporting today and hopefully helps them move the needle on the next step in their own evolution," she said.
