More businesses are beginning to rethink their IT architectures, placing more of an emphasis on decentralization, openness and flexibility.
The release of ChatGPT changed the world. Now, with the dust beginning to settle, businesses are getting a clearer picture of the new data infrastructure landscape. And they’re beginning to rethink their IT architectures, placing more of an emphasis on multivendor, decentralization, openness and flexibility.
The proliferation of AI has forced enterprises to confront long-avoided questions about the control of and dependency on their data, particularly around how reliant many companies have become on a small number of vendors. Some of these key questions include:• “Can I trust a company of a foreign state to provide the intelligence of critical national infrastructure?”• “What if the third-party AI agents no longer have the right skills and I need to change vendors?” The notion of outsourcing infrastructure, business intelligence and operations to a small number of third parties has been a reality of business that now demands scrutiny. Data sovereignty and AI autonomy have become political and boardroom topics, and rightfully so. I’ve spoken with many business leaders who expressed a desire to reverse course on their cloud and vendor strategies to maintain control and flexibility over their most strategic assets—data and AI—and not put all of their eggs in one basket.The vendors and tools that look like winners today might be irrelevant tomorrow. Who’s leading between Microsoft/OpenAI, Google and AWS/Anthropic alone changes by the month; one wrong strategic error could have a material impact on their survival—something Satya Nadella is openly “Therefore, businesses are seeking openness and agility to manage costs and avoid vendor lock-in. They want the ability to choose and change to different technologies for different parts of their business. The industry is responding, with open-source standards like OpenTable formats, Postgres and Kafka APIs breaking down walled gardens. Vendors are being pushed by customers to open up and support interoperability. This shift will also enable organizations to keep data where it’s generated—on-premises, in the cloud or at the edge—or sovereign, without requiring them to sacrifice accessibility or move to a single vendor’s infrastructure. This will also better position businesses for future adaptations and innovations. The leaders of tomorrow will likely be the organizations that prioritize open, flexible and resilient data architectures today. At the core, the goal is to enable AI agents to access data. To work effectively, they need to reach across every layer of the data and technology stack to carry out business processes. Open standards and MCP servers are critical to making this possible. For example, companies like Salesforce, Databricks, Microsoft and Celonis are now offering skilled AI agents to handle business tasks—functioning much like external contractors. But their effectiveness shouldn’t be limited to their own organization’s technology. Open standards such as MCP are needed to unlock broader access. For example, a Salesforce agent and a Celonis agent could collaborate, drawing on real-time data from AWS and Confluent Kafka.What This All Means For Agents Building agent-ready architectures starts by ensuring data infrastructure vendors increasingly adopt open standards and offer federated data access . Generally, technology vendors areBut businesses should take caution: If you don’t trust your human employees with data, you shouldn’t trust AI agents either. Strong data practices—covering quality, governance and developer experience—must be established for people before extending access to agents. The next step is deploying MCP servers to provide agents with data access. Contrary to common belief, these should be developed around specific agentic use cases, which means working closely with vendors to align design and functionality. The shift toward a more open data ecosystem creates new opportunities for businesses with AI agents. Agent-ready data architectures and a multivendor strategy can reduce risk and offer a competitive advantage rather than an inconvenience to businesses.
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Meta AI Confirms Your Data Will Be Used For Ads—Here’s How And WhenMeta has confirmed it will collect interactions with its AI tools to serve targeted ads across Facebook, Instagram and Threads. Here's how, when and who is affected.
Read more »
Sanctioned A7A5 becomes largest non-US dollar stablecoin: DataA7A5, a Russian ruble-backed stablecoin issued in Kyrgyzstan, has emerged as the largest non-US dollar stablecoin despite facing multiple sanctions.
Read more »
Economists are flying blind without key jobs data during the government shutdownFrom Wall Street trading floors to the Federal Reserve to economists sipping coffee in their home offices, the first Friday morning of the month typically brings a quiet hush around 8:30 a.m.
Read more »
Delayed economic data further clouds a hazy economic picture for the Federal ReserveThe shutdown stretching into a second week is forcing Wall Street and the Federal Reserve to fly blind to interpret the strength of the economy as releases of c
Read more »
Victorian Architecture Meets Mid-Century Modernism at This Tranquil Litchfield, Connecticut HotelThe team behind the Hudson Valley’s esteemed Troutbeck Hotel expanded its domain of luxury boutique hotels to Litchfield, Connecticut with Belden House and Mews.
Read more »
U.S. scientist among trio awarded chemistry Nobel for developing new molecular architectureElmira Aliieva is an NBC News intern based in London.
Read more »