WingArc1st has announced DataTetra, a data utilization platform solution designed for enterprise AI use, and began rolling it out on October 21. The platform provides integrated support for everything from data integration and management to AI-powered analysis and business applications. The company says it is designed to address challenges such as rising processing costs as AI use expands and reliance on specific services.
The obstacles to AI adoption go beyond model performance
As more companies incorporate AI into their operations, the obstacle may be less about “which AI to choose” than the mundane question of how to keep working with the data they already have. Corporate data is becoming more diverse and larger in scale, making it more complex to process for use by AI.
One example is “Context Rot,” in which feeding AI large amounts of information makes it harder to grasp the context and leads to less accurate answers. It runs counter to the simple expectation that the more material you feed it, the smarter it will become. Another issue that cannot be overlooked is that a company’s operations can be affected by the availability of external AI services.
Cost is another concern. Processing unstructured data and repeatedly using AI for analysis and search can cause processing costs to accumulate based on usage. The more routinely AI is used, the higher the bill. Some people in charge may recognize the feeling that their wallets are gradually getting lighter in exchange for convenience.
Survey finds about 80% see costs as a challenge
In its September survey, “Survey on Corporate AI Use and Organizational Outcomes,” the company asked whether respondents thought the cost of using AI services would be a challenge for future AI adoption. A total of 79.1% said they “strongly agree” (28.2%) or “somewhat agree” (50.9%).
While adoption is moving ahead rapidly, concerns about costs during the operational phase appear to be widely shared. DataTetra is positioned as a product that directly addresses these concerns.
Bring data together and choose where to use AI
At the heart of DataTetra is a newly offered “lakehouse environment.” A lakehouse combines the strengths of a data warehouse (DWH) and a data lake, allowing structured, organized data and unstructured data such as documents and images to be handled together. The company says the platform supports the entire process, from data integration and management to AI-powered analysis and business applications.
Its three guiding principles are “coexistence with AI,” “data sovereignty,” and “a data utilization platform.”
The idea of limiting what AI is responsible for
“Coexistence with AI” involves creating an environment where AI can refer to the meaning of data and the conditions for its use while working with both structured and unstructured data. The idea is then to combine areas that require AI processing with those better suited to conventional programming, limiting excessive reliance on AI.
This is an interesting point. Rather than expanding the use of AI as an all-purpose tool, routine processing is left to conventional systems, while AI is assigned only the tasks that truly require it. This seems like a sensible division, both in terms of cost and accuracy.
“Data sovereignty” to limit vendor dependence
With “data sovereignty,” companies can build an environment that makes use of data assets they already have, such as a DWH or lakehouse, while limiting dependence on any particular vendor. The aim is “user sovereignty”: enabling companies to choose and control their own data and AI environment.
As a data utilization platform, it provides integrated support for data integration, management and semantic enrichment, as well as building, sharing and using applications and AI agents.
Forecasting costs with capacity limits instead of pay-as-you-go pricing
The solution combines the company’s technologies and products, including Dr.Sum, MotionBoard and dejiren AI. It supports both configurations that complement an existing DWH and those that build a data utilization platform around DataTetra, allowing deployment to suit each company’s system environment.
The pricing model is worth noting. Rather than pay-as-you-go pricing, where customers pay for what they use, the company plans to adopt “capacity-based pricing,” based on usage allocations such as processing capacity. It says the goal is to create an environment where companies can continue using data while keeping costs predictable, even as they expand their use of AI.
That said, the extent to which capacity-based pricing actually reduces costs will likely vary depending on each company’s scale of use and how it uses the service. Still, shifting away from the anxiety of watching the meter run every time AI is used and toward a model that makes it easier to set a budget in advance could give executives and IT departments a practical option to consider.
After deployment, the challenge is embedding it in the organization
Simply installing a platform will not deliver results. The company says it will also offer services to support organizations from building a data utilization environment through operations and embedding it across the organization. Drawing on the expertise of a specialist team that handles more than 100 support projects a year, it plans to help organizations establish practices that sustain data utilization, including staff development and the creation of operational structures.
AI does not deliver value simply by trying it once; its value emerges only when it is used continuously. What matters then is less the intelligence of the model than the foundations that support it: how data is organized, how clearly costs can be forecast, and whether teams on the ground can keep operations running. The announcement of DataTetra reflects how the AI race is shifting from “what to use” to “how to sustain it.”
Comments