Brad Gerstner and Bill Gurley interviewed Satya Nadella, Microsoft's CEO, on their podcast, BG2 Pod. Nadella claimed that the age of SaaS is over and that integrated platforms, with AI as the leading force, are the future.
SaaS applications have powered business operations for decades — seeding the systems that run everything from Customer Relationship Management (CRM) to Human Resources (HR) to Enterprise Resource Planning (ERP) systems and project management applications to collaborative workspaces. Their success has been based on CRUD (Create, Read, Update, Delete) databases with layers of business logic on top. But, with the advent of AI, the entire paradigm is about to change.
With a PhD in Computer Science, Andrea Rosales specialises in domain adaptation, transfer learning, continual learning, and generative AI. Andrea is passionate about developing innovative data science models that deliver impactful solutions. She has a proven track record of creating novel deep-learning models to address real-world problems in both industry and academia, and she is recognised as a Global UK Talent.
Andrea Rosales
Lead Data Scientist
Satya Nadella has described a future where this business logic migrates to an AI layer, transforming the role of traditional SaaS apps. Now, AI agents serve as intelligent orchestrators, integrating multiple platforms and consolidating various workflows, decision-making and automation into a tier of AI.
Essentially, this means that SaaS applications are looking to revert to their most primal form of simple databases. At the same time, the AI layer takes the part of “intelligence”, organising and elaborating processes without being bound to a singular backend. Consequently, SaaS tools will become domain-specific enablers, losing their standalone prominence as the brain of business systems and transitioning fully to AI.
In the following, I will summarise what I took from Nadella's interview and explain the shift from standalone applications to AI agents and what it means for organisations.
Software as a Service (SaaS) is a cloud-based model that delivers software applications over the Internet to the end-user. SaaS products can be used without local installation and are virtually device, location and time agnostic.
The other key feature of SaaS is pricing, which uses a subscription or pay-as-you-use plan. This model permits companies to pay solely for services used and allows for upsizing or downsizing to meet peak seasonal demand.
Saas itself began in the 1960s, when large companies began using software as a service. Yet the idea of modern SaaS, as we know it now, began in the late '90s and early 2000s alongside high-speed internet and cloud computing technology.
During this time, Salesforce pioneered the “no software” model, which set the basis of how software is being used today. Its CRM solution was accessed through a web browser — no software needed to be installed on a local machine. The model was in direct contrast to the software delivery model experienced at the time, which required licenses to be purchased at the top of the hierarchy, while the software updates required manual management.
Gmail, the search giant's cloud-based email service, was launched in 2004. This service showcased the importance of flexibility and access to personal information on the go.
Original Gmail webmail interface. Image taken from Wikipedia.
In 2006, it was Amazon Web Services(AWS) that provided a scalable cloud infrastructure for SaaS solutions to use, enabling businesses to “rent” cloud-based servers and storage. Companies were able to deploy and implement SaaS applications quickly and easily. The service avoided the need to invest in hardware infrastructure.
Launched in 2007, Office 365 transformed traditional productivity software. Here, Microsoft showed us how traditional desktop applications could be evolved into robust SaaS offerings by giving users more powerful updates, more effective collaboration features, and the ability to use these applications virtually.
Dropbox took file and document management to another level since 2011 where users can collaborate and share documents. The cloud storage allows users to query their data on any machine that has an internet connection.
In 2016, Slack was the glue that brought together business communication. It integrated with hundreds of other SaaS applications, all working from a single, centralised productivity centre.
One of the major drivers of growth of SaaS was the COVID-19 pandemic. Cloud computing became a necessity for businesses during the pandemic as the nature of work shifted to remote working.
According to Statista over 50% of companies increased their current SaaS spend in June 2020. SaaS tools were key enablers to keep businesses afloat and these result in exponential growth in video conferencing, collaboration, and project management services. For example, Microsoft Teams kept employees connected via meetings, chats, and calls, providing collaboration, and automated business processes in a single app.
Growth in SaaS app usage. Source: “The 2023 State of SaaSOps”. Image edited by author.
According to the “The 2023 State of SaaSOps” report, organisations currently operate an average of 130 SaaS applications, up from the previous year (2022) by 18%.
The history of SaaS is one of incredible growth and constant innovation. Every era, from mainframes to cloud-based applications, brought great benefits and posed new challenges.
There is no doubt that the SaaS revolution has changed how businesses and consumers engage with technology. Over the last two decades, it introduced organisations to accessing powerful tools in the cloud, reduced infrastructure costs, and instilled scalability within organisations. However, the SaaS model is undergoing a paradigm shift as AI arrives and becomes the backbone for all modern technological advances.
We are entering an era of integrated platforms where AI drives innovation. More and more businesses expect seamless integration across tools, as siloed solutions generate inefficiencies, fragmented data, and disconnected workflows.
According to a report by McKinsey, by 2023, AI will create $13 trillion in economic value. This trend means that AI is no longer just a feature; it's a base layer that will define how software is built and used.
In November 2024, Microsoft introduced the new concept of AI agents, which diverges from traditional software. Unlike traditional standalone applications that execute predefined functions with predefined context, AI agents are designed to process complex, multi-dimensional tasks that used to require multiple applications and manual coordination. For example, rather than a team member extracting sales data from a CRM, generating an Excel report and creating a PowerPoint presentation, an AI agent can seamlessly and independently run all of these steps.
Microsoft called this system generalist agentic systems and introduced Magentic-One. Magentic-One is a multi-agent architecture in which a lead Orchestrator agent guides four other agents to solve tasks like operating a web browser, analysing local files, or writing and executing Python code. It is built on AutoGen, Microsoft's open-source multi-agent framework for developing multi-agent applications.
Example of how Magentic-One operates. Imagen taken from Microsoft's technical report.
Satya Nadella explains how, with Magentic-One, business logic in traditional software will transform to an AI tier where AI agents become the primary orchestrators of workflows and operations. The AI agents don't discriminate between backends or data repositories — they interact with multiple databases seamlessly, functioning as intelligent “multi-repo CRUD” operators.
Nadella emphasises that tools like Microsoft 365 Copilot will act as the organising layer for AI agents in this new era, providing a seamless user experience. In other words, Copilot is the organising layer UI for AI, uniting its own AI agents (for example, Excel or Word as agents) with external systems. This model creates a seamless experience where users can focus on outcomes, and the AI handles the complexity of tool integration and backend management.
In his vision, AI becomes the “brain” of business systems, with SaaS tools evolving into specialised enablers rather than standalone solutions.
This shift means a radical rethinking of how software is implemented and utilised for organisations. In Nadella's view, AI agents will not only orchestrate workflows but also replace traditional business logic currently residing within SaaS applications.
Below, I outline the key implications for organisations based on my interpretation of Nadella's view.
Nadella explains that the backend systems of traditional SaaS tools will become interchangeable. AI agents will interact directly with multiple data repositories and update them seamlessly. This change in the workflow will allow organisations to focus on outcomes rather than tool management.
Intelligent agents act as intermediaries between different software solutions. Rather than jumping between applications to manage customer relationships, analyse data, run marketing campaigns, etc, AI agents will automate and consolidate these operations into one efficient and integrated workflow. This workflow orchestration is a shift from tool-centric processes to goal-centric actions.
The notion of business applications as we know them today will “collapse,” Nadella says. Rather than standalone tools, organisations will find AI as an enabler for decision making and execution. This evolution means businesses must prioritise AI-readiness in their IT infrastructure and workforce.
Nadella exemplifies the shift in traditional workflow with Copilot. AI agents like Copilot will be the central hub for managing all business operations, integrating specialised tools (e.g., Excel, Word) and enabling users to perform complex tasks without leaving the AI interface. For example, Copilot can help draft a legal contract, refine it, and distribute it across platforms with minimal user input.
More importantly, this transition opens up new directions for innovation, helping organisations remain competitive and contend with an ever-changing technological environment.
The SaaS model is being upended, paving the way for AI-driven integrated platforms providing more innovative, efficient, and connected solutions. This breakthrough in autonomous and action-oriented AI agents points toward the future of intelligent automation.
The AI tier shaped by Nadella will allow organisations to achieve new levels of efficiency, innovation, and customer satisfaction. As we enter this brave new world, one thing is evident: AI is not only the future of software but the future of business itself.