In Part One of this series, I covered my February briefing and live demo with Bikash Mohanty, VP of Solution Consulting and Product Strategy at Zycus, on Merlin Intake, Zycus’ AI-powered intake management solution. Read Part One, “The Front Door Problem: A Deep Dive into Zycus Merlin Intake (Part One)” here.
This follow-up offers Ardent’s perspective on what the shift means, why it matters now, and what procurement and AP leaders should be thinking about as Intake-to-Pay strategies continue to develop.
The larger story is not simply AI-enabled request capture. It is the growing realization that procurement spent the last two decades optimizing downstream workflows while leaving the actual point of engagement largely unmanaged. That is beginning to change, and the implications run deeper than most organizations currently appreciate.
Intake Has Graduated
For years, procurement technology vendors treated intake primarily as a user experience problem. Make the form easier. Build a better portal. Add a chatbot. Those improvements mattered at the margins, but they never fundamentally changed organizational behavior, because the problem was never really about the form.
Business users bypass procurement when the process feels disconnected from how work actually happens. In most enterprises, the real front door to procurement is still email, a Teams message, or a conversation in the hallway. Each of those interactions represents spend that never entered the system, a supplier relationship that was never rationalized, a policy that was never applied. AP inherits the consequences downstream in the form of non-PO invoices, mismatched data, and avoidable exceptions that have to be resolved manually.
What the market now calls intake management is evolving into something considerably more significant than a better front-end experience. The more accurate framing is that intake is becoming an operational control layer, the point where governance, visibility, and orchestration either begin or do not begin. Ardent has described this as a procurement control tower: a coordinating layer that sits above the Source-to-Pay process and manages activity from the moment a need is identified through downstream execution. Zycus has built Merlin Intake explicitly around that concept, and based on what Bikash walked us through in February, the product reflects a serious understanding of what that actually requires.
The Architecture Question
The most important question procurement leaders should ask about any intake platform is not whether the interface is conversational or AI-enabled. It is how effectively the intake layer connects requests, policies, workflows, data, and execution across the enterprise. That is fundamentally an architecture question, and it is where the market is beginning to divide in meaningful ways.
Standalone intake providers make a credible argument around flexibility and interoperability. For organizations operating highly fragmented procurement environments or long-established best-of-breed ecosystems, a dedicated intake layer can unify the request experience without forcing broader platform consolidation. That is a legitimate strategy and, for some organizations, the right one.
Suite-native intake models offer a different advantage: operational continuity. When intake, supplier management, contracting, purchasing workflows, invoicing, and analytics operate within a shared architecture, the system has direct access to the operational context surrounding every request. Policy enforcement becomes more accurate. Workflow orchestration runs deeper. Data does not leak between integration points. Downstream processing, including in AP, improves because the information that enters the system at the front end is cleaner, more complete, and more consistently structured by the time it arrives.
Zycus makes the suite-native argument firmly. Bikash described Merlin Intake as native to the broader Zycus S2P platform, with no APIs or integration points between the intake layer and the downstream modules. The operational logic behind that position is sound. The honest caveat is that architecture fit depends on where an organization is starting from. The intake decision and the broader platform strategy are increasingly the same decision, and leaders who treat them separately will likely find themselves revisiting one or the other sooner than expected.
What Zycus Gets Right
Several things stood out in the Bikash briefing that reflect genuine market understanding rather than product marketing.
The first is how conversational AI changes adoption dynamics in ways traditional procurement systems never could. During the demo, a user submitted a catering requisition in plain language, no form, no dropdown menus, no mandatory fields. The platform extracted structured information, identified a supplier onboarding requirement, suggested a category code when the user did not know one, calculated a per-head cost adjustment automatically, and projected the expected PO cycle time based on current policy. The system adapted to the user rather than asking the user to adapt to the system. For organizations that have spent years fighting procurement adoption, that reversal is not a minor UX improvement.
The second is the approach to policy enforcement. Bikash described administrators defining policies in plain language, with the AI interpreting and applying them conversationally during the request process itself, rather than through static rules buried in approval chains that most users never read. Policies that exist outside the actual flow of work are routinely ignored until they create a downstream problem. Embedding guidance into the intake experience means compliance is more likely to happen before the transaction starts moving, not after it has already gone sideways.
The Microsoft Teams integration reflects a practical understanding of enterprise behavior that is easy to underestimate. Procurement technology adoption has always suffered from the same basic problem: users are asked to leave the environment where their work actually happens and go somewhere else. Embedding intake directly into the collaboration platform where employees already spend their time is not a headline AI feature, but it may prove to be one of the more consequential ones in terms of sustained engagement.
The Unresolved Variables
The briefing also surfaced questions that apply well beyond Zycus and matter for any organization evaluating AI-powered procurement platforms.
On AI differentiation: like most ProcureTech providers, Zycus is building portions of its AI capability on top of third-party large language model infrastructure. That is a reasonable and now common approach. But the market has not yet determined where durable AI advantage will actually reside over time. Model access alone will not be sufficient. Workflow intelligence, orchestration depth, proprietary enterprise data, and operational context are likely to matter considerably more than the underlying models themselves as the technology matures. Providers that combine AI capability with deep operational integration are probably better positioned for the long term, but the competitive landscape is still developing.
On data governance: during the briefing, I raised the question directly. How should enterprises think about protecting sourcing intelligence, supplier information, pricing data, and operational knowledge in environments where AI systems are processing that information continuously? This is not a Zycus-specific concern. It applies across the procurement technology market. Organizations becoming more comfortable with enterprise AI deployment should not let that comfort displace scrutiny. Procurement and AP leaders should still be asking detailed questions about data isolation, model governance, security controls, and how proprietary enterprise intelligence is protected over time.
On outcome claims: Zycus cites meaningful reductions in Request-to-PO cycle times and strong compliance improvements tied to Merlin Intake deployments. Those outcomes are plausible, particularly for organizations moving off fragmented or highly manual processes. But technology cannot independently repair weak governance, poor policy alignment, or low organizational adoption. AI can reduce friction and improve orchestration, but process discipline has to exist.
What Procurement (and AP) Leaders Should Be Thinking About
Intake decisions are increasingly architecture decisions. As the intake layer becomes the operational coordination point connecting policy, supplier engagement, spend visibility, and downstream execution, what feels like a platform selection choice is actually a governance strategy choice. Leaders who evaluate intake platforms primarily on interface quality are optimizing for the wrong variable.
AP leaders in particular should be paying close attention to this shift. Many downstream AP problems originate upstream, in incomplete requests, unmanaged suppliers, missing coding structures, and weak policy enforcement at the point of need. Better intake discipline has direct implications for invoice quality, exception rates, touchless processing performance, and supplier onboarding consistency. The connection between what happens at the front door of procurement and what lands on the AP team’s desk is more direct than it often appears, and it is underappreciated in most organizations.
The broader questions worth asking when evaluating any intake platform: How does the intake layer connect to existing workflows, and where does orchestration actually occur? How much operational context is natively available versus dependent on integrations? Will the architecture support broader Intake-to-Pay objectives as those objectives evolve?
The Direction Is Clear
The ProcureTech market spent years solving for the middle and the end of the process while tolerating unmanaged engagement at the beginning. Intake-to-Pay reflects a growing recognition that governance and visibility have to begin earlier, at the moment a need is identified, not after a transaction has already entered the system.
Zycus is not alone in seeing this shift, but the company’s positioning around Merlin Intake reflects a clear-eyed understanding of where operational leverage increasingly sits and what it takes to actually capture it.
What is clear is that the organizations that solve the front-door problem at the right level gain leverage everywhere downstream.
