And we do have a set playbook for Copilot readiness that actually leverages all three suites because people need to prepare, secure and optimize their data state. So step one is prepare is to centralize and enhance data integration integrity and step two is secure, that is identify and also enforce content policies. And last step, optimize is to go forward with governance and automation. You have to do both at the input of the AI model as well as the outputs, because now almost 10% of data generated are being done by Gen Ai. So the output also need to be very much controlled and filtered and moderated because we all know with large language models that you have inherent about 10%, however, around that of hallucination that’s happening. So because output oftentimes is also used as input to data models to capture and model concept drift as the business environment evolves for large enterprises.
So you always have to have that essentially software governance and many in the middle type of approach to ensure that this continuous feedback to the model, it’s managed and measured and optimized. So in fact it’s actually driving all three of our product suites.
Operator: And the next question comes from Gabriela Borges with Goldman Sachs.
Gabriela Borges: Hi, good afternoon. Thank you. TJ AvePoint has this unique vantage point in two ways. One is the savings that you enable for your customers when they think through things like storage optimization, and the other is the projects that you have visibility into from an AI deployment standpoint. So my question for you is how is that push towards cost optimization trending in some of your larger customers relative to last year? And the second part of the question is where is the budget for some of these AI projects coming from and can AvePoint maybe connect the two pieces where you’re enabling savings in one area of the business that then go towards funding AI projects in another area of the business.
Tianyi Jiang: That’s a great question Gabriela. So we continue to see consolidation plays from customers where they like platform vendors, they like less singular products and less vendor, and this decreases risks and allow them to focus on the high quality platform providers. So that economics, focus and platform approach continues into this year. So we do, as you mentioned, give customers significant savings. One of the important value we provide in the Microsoft Cloud ecosystem is to help our customers maximize ROI return on investment on their existing cloud investments. So we ourselves actually consume about $100 million of Azure over a three year period, and that’s growing very rapidly and from that kind of economic scale we’re able to drive further savings for our customers.
So storage optimization is just one such savings. The others are also be able to provide consistent data management capabilities across different license type tiers as well as multi cloud. I can’t emphasize enough that today’s world customers are not only on one hyperscalers. In fact, most enterprise customers, including government agencies, have a mandate to do business continuity reasons. So, they actually intentionally have multiple cloud vendors just to ensure that they have business resiliency. And the second part of your question is the budget bucket. It is true, increasingly the conversations is shifting towards the business budget versus the IT budget. If you only look at pure IT budget, it’s very hard to say, hey, I’m going to spend the extra $30 per user per month just on Gen Ai capabilities.
From an IT infrastructure planning perspective, considering that that’s equivalent to what more than most customers pay for entirety of office we should have today with teams, with Office, with email with one terabyte OneDrive, etcetera combined. However, when you take it into a business context, when you’re driving business outcomes, that it’s a much easier and different conversation and this is where we’re seeing very, very aggressive and active experimentation across the board to drive that business ROIs. So yes, while the Copilot experimentation happening are limited to smaller footprints within the overall user population in companies, but those user populations today are all your power, effectively power users among business community, your head of sales, your head of marketing, head of HR, CFO’s.
So when you have that kind of conversation, it’s a very different conversation than just pure IT, CIO level conversation. So yes, it is actually a very different budget conversation altogether.
Gabriela Borges: That makes sense. And then given some of the dynamics we were just walking through and the newer products that AvePoint has in its portfolio, talk to us about how the go-to-market is evolving, both from a cross-sell standpoint and then from a customer education standpoint as well. How do you intersect some of the problems that you’re solving with what the customers are planning for 2024? How does that conversation get catalyzed?
Tianyi Jiang: Yeah, that’s a great question. So what’s really interesting is AvePoint has been in the business of information management, data management for the last 20 plus years, and it’s a very easy conversation with highly regulated industry customers because that just policy and regulations demand it. But now with Gen Ai rollout and everyone want to take advantage of AI capabilities and disruptions to enhance and innovate on their business, the market’s educating them very, very quickly, the need for a very solid and clean data state. So otherwise it’s a trash and trash issue, because it is not magic. AI models are heavily, heavily rely on your corporate data state. When you do fine tunings on existing large language models to take advantage of these new technology and your existing domain specific industry data, so that your AI capability is not just a summer intern type of knowledge base, but rather 20-year decades worth of industry knowledge in your specific company and domain.
So with that, the awareness of data and information management and governance becomes the forefront across all industries. So that’s really the sea change that we see before we have to educate customers the need for data management and governance and security, and today they’re actually coming to us through ask for that. And as I mentioned with our playbook around what we call oftentimes work with regional partners and Microsoft regional teams around Copilot readiness is that three step process of prepare, secure and optimize and continually optimize and monitor your ongoing data usage to train and refine your models, that actually brings forth all of our product set and our platform. And that also further highlights the need for a platform.
