All views and opinions expressed in this post are my own. Readers should consider this post as a point in time brain dump of my thoughts! – I reserve the right to change my mind at any point in the future! 🙂
Enterprise technology has always continuously evolved in response to an ever changing tempo of business. In the last decade, we saw a huge shift of enterprise change focused on moving to cloud, standardising process and adopting SaaS-based solutions. Though I do question now if those enterprises are operating in a more more volatile environment. Pace of ability to make decisions and spot opportunities to exploit and disrupt the market is not new. But, perhaps what is new (or rather, evolved!) is that that need for pace and competitive advantage across supply chains, financial operations, talent management and customer services may be outpacing what traditional process-centric architectures can deliver. and in reality.. this is probably a bit of a “dog eat dog” situation. If you don’t do it, then someone else will and then you’ll be left behind.
So, if despite the gains achieved across enterprise transformations in the past decade, organisations are still finding themselves stretched across (potentially overcomplex) ecosystems of functions, supplier engagements, geographies, regulatory regimes, geo-political challenges, etc then surely business value is increasingly determined by:
Decision cycle time rather than process cycle time?
Seamless cross-functional coordination rather than siloed or individual workflow optimisation?
Being prepared (and capable) to respond to any variety of change in hours not months (or in some cases, years!)?
Being about to determine what outcome a process, a function or an entire organisational unit has?
Achieving outcomes that genuinely mean something rather than assessing completion rates of specific workflows or processes?
Measuring performance in more innovative ways… like, “did the changes we make actually lead to increased sales and profitability?”
Standing out from the crowd?
Attracting the best and most capable talent?
And to round out this thinking, it would be remis of me not to address the pressures that businesses find themselves in. Finance functions are under pressures to reduce costs, HR functions are asked to respond more dynamically to workforce needs, supply chains must be able to cope with unexpected disruptions and a changing geo-political landscape and public sector must offer more responsive services whilst operating under tighter and tighter constraints. The list could go on and on, and frankly in most cases, the priorities conflict with one another rather than support one another.
So, what’s the answer? If you were hoping to come here and find the meaning of life, I’m afraid you will be sadly mistaken. But in all seriousness, I do have a view. If you agree in full, in part or don’t agree at all, then all outcomes are great, if nothing else, it’s a point of discussion and debate.
Are we in a “Saaspocolypse”?
I know I know. huge clickbait terms like “saaspocolyse” aren’t overly helpful. But they do spark thought and opinion. Forbes (and other media outlets) have been posting a lot lately about this “saaspocolypse”. Here is an excerpt from a Forbes article that you can read here:
“On Tuesday, ~$300 billion of market value evaporated across SaaS, data, and software-heavy investment firms. This was not an earnings miss or a macro shock. It was an AI product release. The damage had been building for months. By the time markets reacted, the IGV Software Index was already down roughly 30 percent from its late-September peak. What changed last week was not the direction, but the speed.
Several of the most entrenched enterprise software companies fell sharply in a single day. Salesforce, ServiceNow, Adobe, and Workday each dropped around 7 percent. Intuit fell nearly 11 percent. At the same time, valuation multiples across the sector compressed violently. The average forward earnings multiple for software companies collapsed from roughly 39x to about 21x in just a few months. Short sellers have already made over $20 billion in 2026 betting against legacy SaaS and are doubling down.
Markets do not erase that much value unless a core assumption breaks. The assumption that just broke was the durability of legacy SaaS growth”
I’m not going to even try and make this post about stocks/shares/market trading/etc. I am as far away from the right person to be speaking about that. but, I will present what I do know and do think and specifically relate that to what I think we can expect from Oracle (specifically, since my expertise are in Oracle) to tackle where the industry may or may not be heading.
What are Oracle doing about it?
For any avoidance of doubt, I do not work for or represent Oracle.
In March 2026, Oracle announced (at Oracle AI World London) a suite of innovations that create a unified, AI-native foundation for outcome-driven enterprise solutions and operations. In my opinion, they signal a new architectural model build around agentic intelligence embedded (in Oracle’s own words) “where work actually happens”. Below I will describe some of those announcements:
Oracle AI Database: Where the data lives
Oracle unveiled a set of enhancements of capabilities within the Oracle AI database through establishing a foundation where AI can operate on trusted enterprise data in real-time. Key innovations include:
Oracle Autonomous AI vector database: which combines the simplicity of a vector database with the full power of Oracle’s AI database and therefore allows developers to build vector-powered applications.
Oracle Unified Memory Core: a unified system that allows AI agents to maintain context across relational, vector, JSON, graph and spatial data within the database itself.
AI Database Private Agent Factory: a no-code environment which allows organisations to build containerised AI agents that operate securely without ever exposing data outside of controlled enterprise environments
Enhanced security layers such as deep data security and Oracle Private AI Containers that are design to defend against emerging AI-era cyber threats such as prompt injection.
Through architecting AI directly into operational databases and data warehousing solutions, Oracle are ultimately eliminating data-movement pipelines which can sometimes be the root cause of AI fragility and governance challenges in organisations (especially highly secure environments).
“Customers don’t just store data – they activate it for AI” – Juan Loaiza, EVP, Oracle Database Technologies
Here is a nice video from Oracle to explain more:
Oracle Fusion Agentic Applications: A system of Outcomes
Perhaps the most consequential announcement (at least in the context of this article) was the introduction of Fusion Agentic Applications. A new category of enteprise application innovation where specialised AI agents can reason, decide and act directly within Oracle Fusion Cloud.
Unlike traditional digital assistants, fusion agentic applications will operate natively inside the transaction engine (rather than on top of) with full access to enterprise data, workflows, policies, hierarchies, audit trails, etc. What this means is, they get the ability to execute complex multi-domain business processes all wrapped up nicely in existing enterprise governance and controls. Oracle describes them as applications that:
Are outcome-driven, operating against explicit business objectives
maintain persisten, shared context across time and steps
Collaborate as teams of specialised agents
Progress work autonomously within agreed guardrails
Surface only decisions where human judgement changes outcomes
As part of the announcement, Oracle launched an initial set of 22 agentic applications spanning finance, HR, supply chain and CX. They include use cases like workforce scheduling, supplier sourcing, sales and cash collection. Ultimately, they are going after high-cognitive-load areas where often a workload or an end-to-end set of processes is distributed across multiple teams (and sometimes systems). Since the initial announcement, more and more agentic applications are becoming available out of the box with quarterly Oracle Fusion releases.
“We are providing applications that can reason, decide and act. We are redefining how work, works.”– Steve Miranda, EVP, Oracle Applications
Since I think this advancement is really neat and will have a genuine impact to how fusion is implemented and used to drive outcomes for organisations, here are a set of videos from Oracle that showcase them:
Oracle AI Agent Studio: Building Enterprise Agentic Applications
To support organisations in tailoring agentic capabilities to their own business need, Oracle have now also expanded their existing AI Agent Studio offering so that it now introduces the following capabilities:
Agentic Applications Builder: this is a natural-language environment that helps organisations to compose outcome-focused applications from Oracle, partner or external AI agents
Workflow Orchestration: for reliable, multi-step and even multi-agent execution
Contextual Memory: that allows AI agents to retain relevant information across many user interactions or steps in a workflow
Content Intelligent: bringing unstructured and structured data together for even more comprehensive reasoning
Observability, Auditability and the Agent ROI Dashboard: this is really key. We hear so often about “failed AI pilots” and this feature gives organisations visibility into performance, risks and the business impacts of their AI investments
What is my view?
It is my view that existing SaaS applications are not insufficient and are not “legacy”.But rather that now is the time for us to start doing something more with them. SaaS solutions like Oracle Fusion Cloud has brought huge benefits to businesses globally and brought into line full standardisation of process and adoption of best pratices.
If organisations now have the standardised (and semi-automated) business processes and the technology to deliver them, now must be the time to bring intelligence, adaptability and continued progression/evolution to the same processes. But, is it worth it? why does it matter?
As I’ve mentioned, for the past 10 years, organisations have lived through a sequence of transformative technology inflection points. They shifted from on-premises solutions to SaaS, adopted the rise of cloud platforms and even started using digital assistant /chatbot capabilities. Each phase delivered progress of course. SaaS, specifically, revolutionised enterprise operations through providing those standardised processes, through elastic scalability, through continuous delivery of enhancements and through governance frameworks that are trusted worldwide across all industries.
These advancements to me now feel like Oracle’s new agentic capabilities extended the very purpose of SaaS. They take the foundational strengths (like security, data integrity, embedded controls, etc) and fuse them together with reasoning and agentic execution. Rather than living as external AI solutions, AI can now be embedded “where work actually happens“:
Inside the data platform, via Oracle AI Database so that agentic AI acts on trusted, real-time enterprise data without needed to move it or duplicate it
Inside the application, via Oracle Fusion Agentic Applications that operate mission-critical business process
Inside the automation layer, via Oracle AI Agent Studio, so that organisations can compose the capabilities that they need to drive business outcomes.
Oracle’s agentic innovations of course do more than introduce new product capabilities. It is my view that for those organisations that adopt Oracle SaaS solutions for their enterprise operation, these innovations fundamentally redesign how organisations will now need to design, implement, operate and measure their enterprise systems. Implementations will become outcome-led rather than process-les and the operating model of existing or future solutions will need to shift too to accommodate new/changed governance controls and hybrid (human and AI) workforces. According to a report written by McKinsey, the use of agentic AI can increase productivity by up to 40% in certain industries – that is definitely going to change how organisations are structured.
And lastly, a concern for many organisations on the enterprise adoption of AI is about risk (especially in public sector where security constraints are high). I hear concerns about data leakage, about hallucinations, about decision errors and about potential compliance gaps. It is also my view that Oracle’s agentic innovation look to address these concerns directly:
Fusion agentic applications operate insideexisting security, approvals and policy frameworks. You already trust your system of record, now just do more with it
Oracle AI Database introduces hardened security capabilities like the deep data security model
AI agent studio includes the enhanced observability and audit logs
This means that in the “new model” (whatever that might become over time), AI becomes safer when it is embedded, not when it is external.
So yes… my perspective is not that (at least in Oracle’s case) we are seeing a rejection of SaaS solutions but rather the fulfilment of the promise that SaaS started with: to deliver technology that continuously improves how work gets done.
Might I be wrong? maybe? maybe not? … I guess time will tell
