The road to agentic artificial intelligence might be paved with stepping stones that progressively construct on one another.
Our research suggests that agentic AI is not going to suddenly appear and not using a strong data foundation built on: 1) cloud-like scalability; 2) a unified metadata model; 3) data mesh organizing principles; 4) harmonized data and business process logic; and an orchestration framework that comes with governance, security and observability.
Though some imagine the yr of agentic AI will come to fruition in 2025, we predict bringing these capabilities together is a decade-long journey and there’s no shortcut on the yellow brick road to realizing agentic automation.
On this Breaking Evaluation, we piece together previous research and point to a dramatic change within the enterprise software stack. We’ll explain how we see the journey playing out, the critical pieces of the emerging enterprise software architecture and the high-value layers of real estate in that system which can be still taking shape.
Background on our research
Over the course of the past two years, we’ve been laying the groundwork for understanding the impact AI has on the enterprise software stack. We’ve tried to chop through the agent washing and highlight the prerequisites for agentic AI success. We’ve discussed the progression of the information stack beyond separating compute from storage. And we’ve emphasized the importance of separating compute from data, underscored by open table formats equivalent to Iceberg and its potential unification with Delta. We’ve also discussed the necessity to unify metadata and the shift of control from the database to the governance layer and the way that piece of the stack is opening up (think Unity and Polaris). This all builds on the work done early on by Zhamak Dehghani with data mesh as an organizational construct for breaking data silos.
Earlier this yr our research focused on configurable business processes in the shape of metadata, using the Salesforce Inc. data cloud as an example. And more recently, harmonizing not only data but additionally shared business logic with examples equivalent to Celonis SE, Palantir Technologies Inc. and RelationalAI Inc.
AI needs a brand new software stack
AI is a catalyst and is disrupting an enterprise software stack that’s five many years old.
Many imagine this AI era is probably the most profound we’ve ever seen in tech. We agree and liken it to mobile’s role in driving on-premises workloads to the cloud and disrupting information technology. But we see this as much more impactful. But for AI agents to work now we have to reinvent the software stack and break down 50 years of silo constructing. The emergence of knowledge lakehouses isn’t the reply as they are only a much bigger siloed asset. Moderately, software as a service as we all know it’s going to be reimagined.
Two distinguished chief executives agree. At Amazon Web Services Inc.’s recent AWS re:Invent conference, we sat down with Amazon.com Inc. CEO Andy Jassy. Here’s what he needed to say in regards to the way forward for SaaS:
I’ll say supply chain is one other area that we expect we may be very effective and now we have a number of experience identical to customer support there. But I also imagine that AI goes to open up all varieties of latest SaaS opportunities and softwares and repair opportunities. I’ve been saying this for a very long time, I’ve told you guys this too, which is that I believe each SaaS company and application that we all know of might be reinvented with what’s available within the cloud. And I believe that’s doubly true when you concentrate on what AI allows.
And Microsoft Corp. CEO Satya Nadella on the BG2 Pod recently went into some depth that we’re going to unpack. Here’s what he said:
The notion that business applications exist. That’s probably where they’ll all collapse, right? Within the agent era, because for those who give it some thought, right, they’re essentially CRUD databases with a bunch of business logic. The business logic is all going to those agents and these agents are going to be multi repo CRUD, right? So that they’re not going to discriminate between what the backend is. They’re going to update multiple databases and all of the logic might be within the AI tier, so to talk. And once the AI tier becomes the place where all of the logic is, then people will start replacing the backend.
Jassy sees cloud plus AI because the transformative catalyst and Nadella talks about multi-repo CRUD databases – which stands for Create, Read, Update and Delete. With the logic within the AI tier, when he talks about replacing the backend, Nadella, like Jassy, envisions a sea change in SaaS.
What Nadella is talking about is basically a 10-year vision without mentioning any intermediate steps on the best way. Someday, we could have the technology to place all of the deterministic rules and logic that constitute an application today right into a nondeterministic neural network, in the shape of an agent. We should not have that technology today. And so, there are various steps between where we’re today and attending to the vision Nadella talked about, and we’re going to undergo those.
What he’s saying, principally, is that we will “kneecap” every SaaS app and switch it into just its database schema. But when we try this, we’ll have one other Tower of Babel, with a bunch of agents that don’t know check with one another — though the vision is that the agents can talk across the databases.
The fashionable cloud data stack is a start line on the journey
Let’s pick up from where we’re today: the present modern data platform. A few years ago we began talking in regards to the sixth data platform beyond the five modern existing data platforms typified by Snowflake and Databricks.
Above is data from Enterprise Technology Research, which shows Net Rating or spending momentum on the vertical axis and Overlap or penetration into an information set of greater than 1,700 IT decision makers on the horizontal plane. We’re plotting Snowflake Inc. and Databricks Inc. together with Google LLC, AWS and Microsoft. We also show Oracle Corp. for context because the legacy database king. The red dotted line at 40% indicates a highly elevated Net Rating. We annotate Microsoft and Oracle because they’re in the information game but they’re not considered representations of the trendy data stack per se. But we don’t wish to debate that today. We show this because these are the players which can be squarely in the combination of this transition. They’ve rather a lot to achieve and far in danger.
The journey on the yellow brick road
As shown below, the cloud data platforms are the start line on our stroll down the yellow brick road.
As reported previously, there may be a shift underway from control on the database layer toward the governance catalog, shown above because the operational metadata. This shift begins to put the muse for a brand new application platform. Horizon from Snowflake and Polaris, its open source catalog, Unity from Databricks, and other established governance platforms equivalent to Informatica, Collibra and Alation are all in play.
Firms are considering in a different way about organizing around data, employing concepts like data mesh and treating data as a product. They’re leveraging details about people, places and things in a distributed organization. The true excitement lies within the movement toward incorporating business processes and harmonizing each data and processes, enabling swarms of agents to work together toward a desired end result.
The unique cloud data platform — Snowflake — was among the many first to separate compute from storage. Over time, the industry has recognized the necessity to separate compute from data. With the rise of open table formats or OTFs, multiple compute engines can access the identical data. This requires separate metadata, including technical and operational details equivalent to lineage. Such metadata formed the muse of knowledge pipelines and created data products defining concepts equivalent to “customer,” “product” or “lead.” Nevertheless, these constructs remain static entities.
To trace a customer’s journey from engagement to prospect, then to guide, and ultimately to conversion, for instance, current methods simulate the underlying business process using business process metadata. This provides a static representation that may be configured per customer, but only to a limited extent.
Salesforce’s Data Cloud stands out for customer data, representing Customer 360 and the complete customer journey in a harmonized way that supports analytics and applications. As a substitute of merely sharing tables, platforms share the concept of a customer and their journey. The following challenge is moving beyond a static metadata picture to sharing the business process logic itself across applications, enabling ultimate flexibility.
With harmonized process logic, agents can communicate across the complete Customer 360 and the shopper journey, using a typical language. Without this, agents must contend with scattered tables. A corporation equivalent to JP Morgan Chase & Co. might need 6,000 tables referring to “customer,” creating a contemporary Tower of Babel that doesn’t function effectively.
This pertains to the notion that the longer term may consist of diverse SaaS applications reduced to schemas, with 1000’s of tables referencing “customer” remaining disconnected. No current technology allows AI agents to harmonize such complexity independently. Symbolic harmonization is required so agents can speak a unified language. This harmonized logic is crucial for achieving true automation. Perhaps within the distant future, it’s going to be possible to discard this logic layer, but for now it stays out of reach.
The shifting points of control and value within the enterprise software stack
Let’s take a have a look at how the enterprise software stack is changing, as highlighted below.
The concept shown above was introduced earlier this yr. At the underside of the stack, AWS represents the underlying cloud infrastructure, setting the stage for others like Google, Microsoft and Oracle (with OCI) to hitch in. Snowflake popularized the separation of compute from storage, essentially providing infinite capability as a cloud data warehouse. Databricks then focused on data science and data pipelines, influencing the shift toward open table formats equivalent to Iceberg. Databricks acquired Tabular and is now working to unify Delta and Iceberg. Amazon’s announcements at re:Invent around S3 tables and open table formats further underscore this trend, aiming for read/write capabilities and governance integration.
The important thing point, highlighted on the left side of the referenced chart, is the shift of control from the database management system to the governance layer. This governance layer is increasingly open source, elevating the importance of what may be termed the “green layer.” This includes the semantic layer, which harmonizes data. Nevertheless, as noted in previous research, the method now goes beyond data — it includes business logic and business processes. That is where the brand new source of competitive value emerges. Salesforce, Palantir, Celonis and others are participating on this evolving ecosystem, making a latest competitive environment.
As previously emphasized, the information platform landscape was once dominated by the DBMS and its control of storage. The opening of the table format meant that the DBMS could not define the state of the tables if other engines were going to read and write to them. Control shifted to the operational catalog. Databricks’ Unity catalog, introduced in 2023, appears to be a powerful contender here. Although there have been statements of direction around open-sourcing Unity that usually are not fully realized yet, Databricks executes rapidly, and the unification of Iceberg and Delta is now expected ahead of we initially anticipated – perhaps as early as Q1 2025.
Snowflake’s Horizon catalog, the brand new source of truth for its ecosystem, still runs atop the Snowflake engine but synchronizes with Polaris. This permits governance policies set in Horizon to be applied to the open Iceberg ecosystem. The following layer up involves adding data semantics for ideas like customers, products, leads, and campaigns — the primary a part of the semantic layer. The far tougher aspect is harmonizing processes, which requires changes to databases which were many years within the making. Achieving it will pave the best way for agents that may operate effectively on this latest environment.
What the software stack looks like in the longer term
Let’s paint an image of what this stack looks like at a gentle state.
The evolution from on-premises environments to the cloud began with infrastructure as a service, which reduced much of the heavy lifting related to infrastructure management. This progression continued with platform as a service and SaaS, where more infrastructure activities — what Amazon calls “undifferentiated heavy lifting” — became managed services. Nevertheless, the green layer at the highest of the stack is where latest value is emerging.
Three layers are shown above within the green: the digital representation of a business, a network of agents, and a brand new layer of analytics guided by top-down organizational goals. This structure enables the interpretation of goals and adjustments based on market changes or human guidance. The result’s bottom-up outcomes driven by agents collaborating with one another and with humans, while taking motion in a governed manner.
This latest set of layers integrates the silos of applications and data built over the past 50 years. These silos can now be abstracted and become what Nadella described as “sediment.” Nadella’s viewpoint focuses on the information layer, while this angle emphasizes the applying logic layer that hosts the agents.
There may be a transparent business imperative behind this shift. We imagine corporations will differentiate themselves by aligning end-to-end operations with a unified set of plans — from three-year strategic assumptions about demand to real-time, minute-by-minute decisions, equivalent to pick, pack and ship individual orders to fulfill long-term goals. The function of management has at all times involved planning and resource allocation across various timescales and geographies, but previously there was no software able to executing on these plans seamlessly across each time horizon.
This end-to-end integration requires a harmonized digital representation of the business as a foundation. With this, analytics can orchestrate and align agent activity that happens not only inside silos but additionally in collaboration with humans. Management thus becomes increasingly integrated right into a software system — an evergreen capital project that isn’t truly finished. As a substitute of relying solely on tacit knowledge stored within the minds of a management team, this data is steadily converted into an ever more integrated software product.
Latest high-value layers are emerging within the stack
Let’s zoom in a bit on a few of the high-value pieces of the stack that we’ve highlighted previously but are value reviewing.
The evolution we envision connects backend systems — each analytic and operational — to extract business logic previously trapped inside applications and make it more accessible in real time. Moderately than relying solely on analytic systems that produce historical snapshots, this approach goals to enable continuous decision making and automating workflows. Two layers stand out, highlighted in red:
- A unifying layer that harmonizes data and business logic.
- An agent control framework that orchestrates and communicates across agents and with humans.
At the highest, organizational goals guide the method. A high-level goal, equivalent to gaining market share, may set constraints around margins or pricing and specify revenue targets and tactics to attain an end result. Agents can understand these goals and execute bottom-up actions inside defined guidelines. Working along with other agents and with human input, these “employee bee agents” adjust to changes available in the market and cling to top-down frameworks.
That is critical since the metric tree — representing business goals from forward-looking strategies at the highest to more technical and operational states at the underside — isn’t only a set of dashboards or historical reports. As a substitute, these metrics function like dials on a management system. Relationships between them have to be learned over time. By applying predictive and process-centric platforms, organizations can conduct experiments, observe outcomes and refine their understanding of how market demand shaping or other actions influence results.
When integrated with models and training cycles, agents learn from each human interventions and observed outcomes. If an agent encounters an exception it cannot handle, a human can step in to guide the resolution. Over time, the agent learns from these “teachable moments” and may handle similar situations independently. Likewise, when agents try and shape demand and measure the consequences on metrics, they gain deeper insights that improve their future performance.
This learning framework — harmonized data, business processes and metric-driven goals — offers a scarce and highly invaluable layer within the enterprise stack. Though there could also be many agents, there might be relatively few such business process platforms inside any given organization. Ultimately, as agents learn from each direct human intervention and the outcomes of their actions, they improve constantly, driving innovation and operational efficiency.
Which technology vendors are leading the solution to agentic?
There are a lot of participants, but below are a few of the players that we’re tracking on this latest world and where we see their value-add within the stack.
The underside layer includes data platform providers equivalent to Snowflake and Databricks, that are leading efforts to represent core business entities. Other corporations like Relational AI, Celonis and EnterpriseWeb LLC are constructing cross-silo capabilities, sometimes called the business process layer. Above that layer, organizations equivalent to Palantir, Oracle and Salesforce are harmonizing business processes inside their very own ecosystems. Moving further up, an agentic orchestration layer is emerging, featuring corporations like Google, Microsoft and UiPath Inc. It’s widely anticipated that AWS can even play a major role on this evolving stack, based on recent announcements and developments.
A key point is that it is much tougher to maneuver from representing business entities — people, places and things — to defining and aligning cross-silo business processes. The industry has spent many years constructing the information and application logic technologies needed to fuse these elements together.
Relational AI, for instance, uses a relational knowledge graph, allowing organizations to declaratively define application logic, much like expressing requirements in SQL. This dramatically simplifies the strategy of articulating logic. Celonis provides business process constructing blocks in order that customers can conduct process mining and configuration with minimal coding. Palantir excels at connecting deeply into core transactional systems but requires more procedural coding, because it doesn’t supply out-of-the-box application templates. Salesforce, with its Data Cloud, offers comprehensive coverage of the complete customer 360 domain, including customer journeys and touch points, expressed through configurable business logic that matches its model. UiPath is able to automate processes, including those where APIs will not be available.
These approaches highlight the complexity of harmonizing business logic across multiple platforms. Constructing a metrics tree of business outcomes requires a consistent representation of enterprise processes. This goes beyond simply connecting schemas from various applications. The metrics tree represents the “physics” of a business — its behavior and logic — linking high-level goals equivalent to gaining market share to more granular operational metrics. Without harmonizing the underlying application logic, it’s difficult to create this cohesive representation of business outcomes.
Briefly, though corporations have made substantial progress in harmonizing data at scale, the subsequent frontier involves fully integrating each data and business processes right into a unified stack. Achieving it will unlock the potential of agentic orchestration and deliver a brand new level of automation, insight, and adaptableness for the enterprise.
How AI will achieve 10X productivity impacts
The large opportunity ahead was explained by Erik Brynjolfsson within the graphic below, annotated by George Gilbert.
The frenzy around enterprise AI largely pertains to boosting productivity. On the surface, this implies achieving similar or greater outcomes with fewer employees. Industry observers, equivalent to David Floyer, often discuss realizing the identical results with a small fraction of the workforce. The important thing query is how it will play out inside the enterprise.
Erik Brynjolfsson’s perspective, depicted in an influence law curve above, is helpful here. Historically, packaged applications addressed certain high-volume, repeatable business functions — often within the back office and other well-defined domains. Custom modifications and applications were then introduced to handle proprietary processes, specialized data, vertical industry tasks or unique organizational needs. These encompassed one other sizable portion of automation.
Yet beyond these implementations lies a really long tail of workflows that remain unautomated. This long tail represents the space where AI agents can deliver an order-of-magnitude increase in productivity. Moderately than relying solely on precoded, deterministic logic — as seen in traditional packaged software — the subsequent generation of agents will learn dynamically. They may adapt to unanticipated scenarios and exceptions by observing outcomes, incorporating human feedback and refining their responses over time.
In other words, on the left end of the curve, tasks were automated through deterministic logic because they were well-understood and repetitive. Further along the tail, there are countless less-common, more nuanced workflows that can not be fully predefined. By deploying AI agents that learn and improve constantly, enterprises can steadily tame these unstructured, long-tail processes. Nevertheless, current technology cannot simply discard existing deterministic rules and rely entirely on a mess of autonomous agents. Doing so would end in a chaotic environment — effectively back to a Tower of Babel — where agents struggle to grasp their roles and responsibilities.
The trail forward involves fastidiously combining traditional, deterministic systems with learning agents, enabling them to handle each well-understood tasks and emerging, unpredictable scenarios. Over time, as agents learn from outcomes and human intervention, more workflows may be automated, significantly increasing overall productivity.
2025 outlook and the longer term of agentic AI
Let’s end with a sit up for 2025 and beyond.
As previously discussed, there may be prone to be significant “agent washing” in 2025. Many will market single agents or lightweight solutions as agentic systems, but closer inspection will reveal that the journey is barely starting. Some claim agentic AI capabilities might be widespread next yr, yet substantial work stays. This isn’t a short-term trend; it is predicted to be a multiyear process. Though efforts may begin to take shape in earnest in 2025, the true impact may take two to 10 years to totally unfold.
One major concern is that vendor-specific agents will emerge inside application silos, reinforcing fragmentation and risking one more unfulfilled promise by the tech industry. Many previous initiatives — equivalent to Customer 365, certain data warehousing efforts and the big-data craze — failed to fulfill lofty expectations. Although the cloud has largely delivered on performance, plenty of data-related guarantees have been broken or only partially realized. The danger is that the industry may simply bolt agents onto existing legacy architectures, effectively “paving the cow paths” quite than delivering a meaningful transformation.
The chance is to reinvent the applying stack quite than perpetuate the establishment. Though leaders equivalent to Jassy and Nadella have acknowledged the necessity for change, even they concede there are challenges and uncertainties in how it will develop. The vision of an agentic future that delivers a 10x productivity gain hinges on harmonizing end-to-end business processes, ensuring that agents and humans collaborate effectively and share a typical understanding.
Different vendors are placing varied bets. Major cloud providers are setting forth their strategies, data platforms like Snowflake and Databricks are staking their positions, and a various group of application players — including ServiceNow Inc., Salesforce, Oracle and SAP SE — are shaping their very own approaches.
Meanwhile, a flood of investment is pouring into agent startups. The query is whether or not these emerging players will help integrate the stack or create latest silos. A lot of these agents will need access to the business logic currently locked inside existing applications. Without harmonized logic and accessible platforms, these agents could struggle to deliver meaningful value, forcing them to pay a premium to tap into that logic.
This underscores the importance of not skipping crucial steps. The trail forward involves creating latest infrastructure layers, incorporating real harmonization and avoiding the trap of superficial bolt-ons. Although realizing this vision will take time and persistence, specializing in the pieces that don’t yet exist — and making them real — can substantially increase the likelihood of achieving the productivity gains envisioned by this agentic era.
What do you’re thinking that? How are you eager about agents in your organization? What steps are you taking to arrange?
Tell us.
Image: Amila Vector/Adobe Stock
Disclaimer: All statements made regarding corporations or securities are strictly beliefs, points of view and opinions held by SiliconANGLE Media, Enterprise Technology Research, other guests on theCUBE and guest writers. Such statements usually are not recommendations by these individuals to purchase, sell or hold any security. The content presented doesn’t constitute investment advice and mustn’t be used as the premise for any investment decision. You and only you might be answerable for your investment decisions.
Disclosure: Lots of the corporations cited in Breaking Evaluation are sponsors of theCUBE and/or clients of Wikibon. None of those firms or other corporations have any editorial control over or advanced viewing of what’s published in Breaking Evaluation.
Your vote of support is very important to us and it helps us keep the content FREE.
One click below supports our mission to supply free, deep, and relevant content.
Join our community on YouTube
Join the community that features greater than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and lots of more luminaries and experts.
THANK YOU
