The Future of Enterprise Applications Consulting in the Age of AI
There is a useful exercise when thinking about the impact of AI on any industry: ask not what the technology can do, but what the economics demand.
The enterprise applications consulting industry is built on a simple economic model. Vendors build complex software. Customers lack the expertise to implement it. Consultants fill the gap.
This model has generated hundreds of billions of dollars across SAP, Oracle, Salesforce, Microsoft, Amdocs, Infor, and every other major platform vendor over the past three decades.
The reason this model worked is that enterprise software was genuinely difficult. Configuring an ERP system required deep, specialized knowledge that took years to acquire. The supply of that knowledge was limited. The demand was constant. And so consulting rates stayed high, SI firms scaled on headcount, and functional consultants with ten years of experience in a single module could command premium day rates indefinitely.
AI changes the supply side of that equation. And when the supply side changes, everything else follows.
What AI Actually Does to Consulting
The mistake most people make when analyzing AI’s impact on professional services is focusing on the task level. “Can AI write a functional specification?” Yes. “Can AI generate a data migration script?” Yes. “Can AI build an integration mapping?” Increasingly, yes. But the more important question is what this does to the structure of the industry.
Consider the functional consultant. Today, a significant portion of their value comes from knowing how the system works. Where the configuration lives. How the setup screens behave. What the dependencies are between modules. This is knowledge that is difficult for a human to acquire but trivially easy for an AI to index.
This does not mean functional consultants disappear tomorrow. It means the price a client is willing to pay for that knowledge compresses. Rapidly. Because the alternative is no longer “hire another consultant.” The alternative is “ask the AI.” And the AI is available immediately, at near-zero marginal cost, around the clock.
The same logic applies to technical consultants writing reports, integrations, and extensions. It applies to project managers whose primary output is status consolidation. It applies to any role where the core value is information retrieval and execution rather than judgment.
What AI cannot do, and this is the critical distinction, is exercise judgment across ambiguous, high-stakes business decisions. It cannot sit in a room with a CFO and a CIO who disagree on the target operating model and navigate that conversation toward a resolution. It cannot look at a client’s organizational structure and know that the implementation will fail unless the org design changes first. It cannot assess whether a vendor’s roadmap claim is credible or whether a system integrator’s estimate is realistic. Judgment, by definition, requires accountability. And accountability requires a human.
The Impact by Role
Functional Consultants
Functional Consultants are the most exposed. The configuration knowledge that took a decade to build can now be replicated by AI trained on product documentation, community forums, and implementation patterns. The functional consultants who survive will be those who moved beyond configuration years ago into business process design, industry-specific advisory, and solution architecture. The ones who stayed in the configuration layer have two to four years before the market reprices their work significantly. Their path forward is toward the client side, into product ownership and business process authority roles, or upstream into advisory work where the system is one input among many.
Technical Consultants
Technical Consultants face a slightly longer runway but the same trajectory. Code generation is improving at a pace that makes mid-level development work increasingly automatable.
Within three to five years, the market for consultants whose primary skill is writing ABAP, PL/SQL, or integration middleware logic will contract meaningfully. The survivors will be those who think in systems, not syntax. Enterprise architects. Integration architects. Platform strategists. The people who decide what should be built and why, not the ones who write the code.
System Integrators
System Integrators face the most structural threat because AI attacks the foundation of their business model. SIs scale by adding people. AI reduces the number of people needed. This is not a future scenario. It is already visible in how leading firms are restructuring delivery teams and investing in automation accelerators. The SIs that survive will be those that shift from selling time to selling outcomes. That build proprietary AI tools for delivery. That develop genuine industry expertise rather than generic implementation capability. The rest will compete on price in a declining market. Consolidation is inevitable. Within five years, the mid-tier SI landscape will look fundamentally different.
Customers
Customers are the primary beneficiaries. For decades, customers have been dependent on their SI for system knowledge. AI breaks that dependency. Clients will be able to interrogate their own systems, validate what their consultants tell them, generate their own test cases, and understand their own data without waiting for a support ticket to be resolved. This does not eliminate the need for external expertise. It raises the bar for what qualifies as expertise worth paying for.
OEMs and Platform Vendors
OEMs and Platform Vendors will capture an increasing share of the post-implementation value chain. As SAP, Oracle, Salesforce, and others embed AI natively into their products, the product itself becomes more self-sufficient. More self-configuring. More self-healing. This is good for customers. It is problematic for the consulting ecosystem that exists between the vendor and the client. The space in which third-party consultants operate is being compressed from both sides. Vendors from above, AI tools from below.
Project Managers
Project Managers will bifurcate sharply. The reporting-focused PM, the one whose primary output is a consolidated status deck and a RAID log, will be automated within three to five years. The delivery-focused PM, the one who manages stakeholder dynamics, makes trade-off decisions under pressure, and leads programs through ambiguity, will become more valuable. There is no AI substitute for the person who can walk into a failing program and make the difficult calls required to get it back on track.
The Timeline
Here is where most analysis on this topic becomes vague. So let me be specific.
By 2028, the majority of standard functional configuration work across major ERP, CRM, and HCM platforms will be AI-assisted to the point where it requires half the consulting effort it does today. Not zero. Half. That alone restructures the economics of every SI engagement.
By 2030, code-level technical work for standard integrations, reports, and extensions will follow the same trajectory. Custom development will still exist but the volume of work that justifies dedicated technical consultants on a project will shrink considerably.
By 2032, the SI industry will have consolidated significantly. The firms that remain will operate with fundamentally different delivery models. Smaller teams. Higher expertise per person. AI-augmented throughout. The era of the 200-person implementation team will be over.
What will not change, and this is the part most people overlook, is the need for people who can make sense of complexity. Who can look at a business with fifteen entities, four regulatory jurisdictions, and a legacy landscape of thirty integrated systems and determine the right path forward. That is not configuration. That is not code. That is judgment. And the market for judgment will grow as everything else compresses.
So What Should You Do
If you are a consultant
Stop investing in platform-specific configuration knowledge as your primary asset.Start investing in understanding industries, business models, and organizational dynamics. The system is becoming easier to configure. The business is not becoming easier to understand.
If you are an SI leader
Look honestly at your delivery model and ask what percentage of your revenue comes from work that AI will be able to do in three years. Then build a plan for what replaces it. If you do not have that plan today, you are behind.
If you are a customer
Recognize that the leverage in the consulting relationship is shifting toward you. Use it. But also recognize that cheaper consulting is not the same as better outcomes. The advisor who challenges your assumptions is worth more than the one who configures your system at a discount.
If you are a platform vendor
Understand that every AI feature you embed into your product reduces the need for the partner ecosystem you depend on for distribution. That tension will define vendor-partner relationships for the next decade.
The consulting industry as it exists today was built for a world where enterprise software knowledge was scarce and expensive. AI is making it abundant and cheap. The professionals and firms that understand this will build the next era. The ones that don’t will spend the next few years optimizing a model that no longer works.
The window is open. It will not stay open long.