AI tooling works best when it has access to relevant business data. Snowflake CEO Sridhar Ramaswamy argues that business applications run the risk of simply being a ‘dumb data pipe’ for the ‘big brain’ of AI. Is this a valid concern?
There is no doubt that this fear exists. Dark red stock prices for Salesforce, HubSpot, ServiceNow, and many other software companies over the past few months suggest that investors are anything but reassured. Although every IT vendor has an AI-related story ready and can be an agentic management platform, Ramaswamy sees that this is not the approach of AI vendors.
According to the Snowflake CEO, the model makers want to create a world in which all data is easily accessible. The applications in which that data is used are therefore essentially ballast. Why use a built-in AI assistant in your ERP or CRM platform when you can access the same data via a central intelligence?
Freedom of choice
Ramaswamy’s warning carries weight. As a central data platform, Snowflake, like Databricks, is an extremely useful partner for AI players. Snowflake recently entered into a $200 million partnership with OpenAI. Models such as GPT-5.2 will therefore be made directly available within Snowflake Cortex AI.
In this way, Snowflake avoids being a ‘dumb data pipe’ for an LLM. But Ramaswamy argues that his company lives with the fear that organizations will stop using AI agents built by software vendors. There must certainly be added value for these specialized agents, for example, that they are more accurate, operate more securely, and are easier to use. For experienced users of existing platforms, this is already the case. A solution such as NetSuite or Salesforce offers AI functionality as an extension of familiar systems, whereby adoption of these features almost always takes place without migration.
Ramaswamy believes that customers have the final say on this. If they want to consult a central AI and ignore traditional enterprise apps, then they should be given that option, according to the Snowflake CEO. The implicit message: software vendors should simply build in the functionality that is requested, or that is so attractive that the alternative pales in comparison.
Simplicity is incredibly complex
The defense we get from software players is now familiar. Organizations are not greenfield operations and cannot simply replace their business suite ‘with AI’. Processes and policies have been formed around the current IT stack, and a migration from that supposed legacy ends up in a world that has to transfer not only the data, but also the daily business of companies.
However, the tug-of-war around the center of AI is in full swing. It is not without reason that vendors claim that their solution should be the central AI system, for example because they contain enormous amounts of data or because they are the most critical application for certain departments. So far, AI trends among these vendors have revolved around the adoption of AI chatbots, easy-to-set-up or ready-made agentic workflows, and automatic document generation. During several IT events over the past year, attendees toyed with the idea that old interfaces may disappear because every employee will be talking to the data via AI.
That simplicity is incredibly complex, because checking all the details remains necessary when using AI. Sometimes tabs are needed to figure out exactly what the data is saying. If this is just data in a “dumb data pipe,” it suggests that the data actually functions as a stream or water, a substance that needs to go from A to B, with B in this case being the AI tool. But that data is not so unambiguous that an AI system can handle it in a single way. Screens within business applications do more than just represent the data as the truth. People also have to interpret that data, and if they only do so through the abstraction layer of AI, this is already guided and focused on convenience. This simplification of data will rub many people the wrong way.
Read also: Oracle NetSuite introduces the missing piece of the puzzle for its platform