Procright - How Orchestrated Agents Transform Procurement

Why Procurement’s Next Leap Is Not Another Suite, but Better Orchestration

Procurement teams have spent years trying to tame fragmented systems, inconsistent intake channels, and slow handoffs between people, platforms, and policies. The promise of "digital procurement" often delivered another interface, another module, or another workflow layered on top of an already complex environment.

What emerged in this discussion is a more practical view of the future: procurement transformation will not come from adding more software for its own sake, but from orchestrating work across systems, data, and increasingly, AI agents.

The conversation brought together three perspectives that matter to enterprise buyers: a procurement practitioner leading innovation at Pfizer, an industry analyst studying intake and orchestration, and a technology executive focused on customer value. Together, they outlined where orchestration is maturing, where AI hype still obscures reality, and how enterprises should evaluate the next generation of procurement operating models.

For procurement leaders, the message is clear: intake may now be expected, but orchestration is where real differentiation and business value are being created.

Key Takeaways

Treat intake as a starting point, not the end state. The real value lies in orchestrating work across approvals, supplier onboarding, sourcing, contracts, and adjacent functions.
Build the business case around a painful, measurable process first. Fast wins in areas like approvals or supplier onboarding can justify broader transformation.
Evaluate orchestration providers with live proof, not polished demos. Test with your data, your integrations, and your edge cases.
Ask whether AI is actually taking action. If it only summarizes or suggests, it may be helpful - but it is not necessarily agentic.
Design for governance before scale. As teams start creating agents, organizations need a way to promote, monitor, and standardize what works.
Be explicit about why you need orchestration. Your use case, process constraints, and integration needs should drive vendor selection.
Expect the market to change. Mergers, adjacent expansion, and cross-functional use cases will reshape the category over the next few years.
Use orchestration to revisit make-versus-buy decisions. Some capabilities may no longer require a large standalone system if they can be built more precisely in the orchestration layer.
Think beyond procurement. Strong orchestration capabilities can eventually support legal, compliance, finance, and IT workflows.

Intake Is No Longer the Differentiator

One of the strongest insights from the discussion was that intake has become table stakes.

That may sound surprising given how much attention intake still gets in procurement technology marketing. But the analyst perspective in the conversation was straightforward: most enterprise platforms, whether legacy suites or specialist tools, already provide some form of intake. The real question is no longer whether a user can submit a request through a front door. It is whether the organization can ensure that request is handled intelligently, consistently, and efficiently after submission.

That distinction matters.

A strong intake layer helps users start the process. A strong orchestration layer ensures the process actually gets completed - with the right routing, decision logic, data dependencies, approvals, and downstream actions across multiple systems.

In practical terms:

Intake captures demand
Orchestration coordinates fulfillment
Agentic automation begins to execute pieces of fulfillment autonomously

This is an important shift for buyers. If procurement leaders continue to evaluate vendors primarily on user interface or front-end form design, they may miss the more strategic question: how well can this platform coordinate work in a messy, real enterprise environment?

Why Orchestration Is Harder Than It Looks

The panel repeatedly returned to a point many procurement teams learn too late: orchestration appears simpler from the outside than it is in practice.

On the surface, orchestration can look like a combination of workflow automation and system integration. That simplicity is deceptive. In real enterprises, orchestration must account for:

Multiple ERPs and source-to-pay tools
Inconsistent master data
Variations in approval logic by category, region, or business unit
Supplier record complexity
Exception handling
Evolving compliance requirements
Organizational nuance that is often undocumented

This is why the analyst on the panel argued that building true enterprise-grade orchestration internally is rare and difficult. Organizations can often build a basic intake form or a lightweight routing workflow. But sustaining robust orchestration over time - especially when integrations fail, requirements change, or edge cases multiply - is another matter.

That observation aligns with a broader enterprise software reality: the hardest part is not launching automation, but maintaining resilience when conditions change.

For technical decision-makers, this means vendor evaluation should focus less on ideal-state diagrams and more on operational durability:

How does the platform handle integration failures?
How quickly can workflows be adjusted?
How transparent is exception handling?
What underlying data model supports decision-making?
Can the platform adapt to different process semantics across business units?

These are not cosmetic concerns. They determine whether orchestration reduces complexity or simply hides it temporarily.

Pfizer’s Case: Start Small, Prove Value, Then Expand

The most concrete operational lesson came from Pfizer’s approach.

Rather than beginning with a massive end-to-end redesign, the team pursued three workstreams in parallel to generate near-term value and justify investment. According to the discussion, the early focus included approvals, supplier onboarding, and intake for selected procurement processes such as sourcing and contracting.

What is notable here is not just the technology choice, but the sequencing.

Pfizer did not begin by trying to solve every procurement problem at once. It began with processes that were:

Painful enough to matter
Defined enough to operationalize quickly
Measurable enough to support a business case

That approach reflects a mature transformation principle: enterprise innovation moves faster when it is attached to a specific operational pain point and a credible return story.

For procurement leaders under budget pressure, this is likely the most transferable lesson from the discussion. If your CFO or transformation office is skeptical, do not lead with abstract claims about AI, orchestration, or digital transformation. Lead with one process where:

Cycle time is clearly excessive
Manual effort is obvious
User frustration is visible
Cost of delay or rework can be estimated

Then prove it.

The Pfizer example also suggests another useful discipline: use early wins to buy time for better design. A quick-return deployment can create organizational confidence while the team learns where the supposedly "well-defined" process is less stable than expected. That is often the hidden payoff of early orchestration work - it reveals the gap between process documentation and operational reality.

AI in Procurement: Move Past the Label and Ask What It Does

The panel’s discussion of AI was especially useful because it avoided both cynicism and hype.

The analyst offered a practical test for whether something is meaningfully agentic: is it taking an action that a human would otherwise need to perform? That is not a complete technical definition, but it is a highly effective buying question.

This matters because many procurement technologies now present AI-enhanced features that range from genuinely autonomous execution to little more than improved assistance. Those are not equivalent.

A useful framework from the discussion is to distinguish between three levels of AI application:

1. Assistive AI

This includes summarization, recommendations, drafting, or question answering.

These tools can improve productivity, but they still rely on human follow-through.

2. Deterministic agentic behavior

This involves AI or rule-driven logic executing predefined steps based on clear instructions.

The panel suggested that some current "agents" function more like simplified robotic process automation with more natural interfaces. That does not reduce their value. In many procurement scenarios, predictability is an advantage.

3. Autonomous agentic behavior

This is where an agent exercises some degree of judgment or decision latitude.

That opens up greater potential - but also greater risk. Procurement leaders must be careful with use cases involving compliance, approvals, external commitments, and supplier-facing actions.

The practical insight here is not that one model is better than the other. It is that different procurement tasks require different levels of autonomy.

For example:

Updating structured supplier information may tolerate deterministic automation
Drafting a sourcing event could benefit from assistive AI
Negotiation strategy or exception-based policy decisions may require human review even if AI contributes analysis

The buying implication is simple: do not ask whether a vendor has agents; ask what the agents are allowed to do, how they are governed, and how performance is monitored.

The Emerging Operating Model: Data, Experience, Process

One of the strongest practitioner insights in the discussion was the idea that procurement innovation increasingly depends on bringing together data, user experience, and process in one place.

That triad deserves attention because it captures why so many transformation efforts stall.

Many procurement organizations have:

Data without usability
Workflow without context
User-facing tools without process depth
AI pilots without operational integration

Orchestration becomes powerful when it connects these elements rather than treating them separately.

In the Pfizer example, the value was not just automation. It was that the platform made experimentation more accessible to people who were not deeply technical. Once workflows, data integrations, and task structures became visible in one environment, business users could begin identifying where agents or automation might help.

That is a significant cultural shift.

Instead of innovation being concentrated in a small technical team, orchestration can create a shared language for process improvement. When business users start saying, "that task could be handled by an agent", the organization has moved beyond passive software adoption and into active redesign.

For enterprise leaders, this is a subtle but important metric of progress: not just whether a platform is live, but whether it changes how teams think about work.

Why Vendor Evaluation Should Be More Demanding Than Usual

One of the most valuable parts of the discussion concerned vendor selection.

Pfizer’s team reportedly required providers to prove their claims in a live context using the company’s own systems and data. That is a much stronger evaluation method than a conventional scripted demo, and it aligns well with the realities of orchestration.

In categories like orchestration, the gap between demo performance and enterprise performance can be large. Demo environments are curated. Real environments contain broken fields, unexpected dependencies, security constraints, regional variants, and integration issues that expose the platform’s true strengths and weaknesses.

For procurement and IT leaders, a more rigorous evaluation process should include questions like:

Can the provider configure a real use case quickly?

Speed-to-value claims should be tested under realistic conditions.

How does the provider respond when something fails?

The discussion highlighted a key truth: problems will occur. Responsiveness and adaptability matter as much as initial functionality.

Does the provider understand your process semantics?

A vendor may know procurement in general but still miss how your organization defines suppliers, approval objects, risk checkpoints, or contract triggers.

What is configurable by the customer versus dependent on the vendor?

This becomes especially important when processes evolve or agents need to be adjusted.

Is the architecture open enough for future interoperability?

That includes integrations not only with source-to-pay systems, but potentially with internally built agents or adjacent enterprise workflows.

This style of evaluation is particularly important in a market where messaging often outpaces operational maturity. Buyers should reward proof, not positioning.

Make, Buy, or Orchestrate? The Decision Framework Is Changing

A notable theme from the Pfizer perspective was that orchestration is changing how enterprises think about the classic make-versus-buy question.

Historically, procurement technology decisions often revolved around two poles:

Buy a suite and accept its constraints
Buy best-of-breed tools and manage the complexity

Orchestration introduces a third option: buy selectively, then compose capabilities more flexibly through a coordinating layer.

That can alter investment logic in several ways.

An organization may decide that it does not need a large standalone tool for every capability if a lighter solution - or even a custom-built function within an orchestration platform - solves the actual business requirement. Conversely, it may still choose a specialist product where depth truly matters.

This is not an argument against suites or against best-of-breed tools. It is an argument for greater precision.

Too many enterprises buy broad capability sets they never fully use. Orchestration can help expose what is actually required, what can be simplified, and where expensive functionality is not justified.

For procurement leaders, that means future architecture decisions should be grounded in three questions:

What business problem are we truly solving?
What level of functionality do we actually need today?
Can orchestration bridge gaps more efficiently than another major purchase?

This more modular mindset may become one of orchestration’s biggest strategic benefits.

The Market Will Not Stay Still

The discussion made clear that the intake and orchestration market is in motion.

That instability is not necessarily a negative. In fact, it is typical of categories moving from early adoption toward broader enterprise relevance. The panel referenced ongoing mergers, partnerships, and directional shifts among vendors. Some orchestration providers are expanding toward broader source-to-pay capabilities. Some larger suite vendors are moving to incorporate orchestration more directly. Others may push into adjacent enterprise functions.

The analyst’s view was especially useful here: while the exact market structure several years from now is not specified in the video, it is unlikely to look like today’s market.

That has strategic implications for buyers.

When the category is fluid, vendor selection should not focus only on current features. It should also consider:

Product direction
Ecosystem openness
Extensibility
Cross-functional potential
Risk of lock-in
Vendor responsiveness to market changes

In a dynamic market, a rigid platform can become a strategic constraint. An adaptable one can become a leverage point.

The Most Interesting Future: Beyond Procurement

Perhaps the most consequential idea in the conversation was that orchestration may not stop at procurement.

This is where the category becomes more than a process-improvement tool. If orchestration is fundamentally about connecting requests, systems, people, and actions to get work done, then its relevance extends into any function with fragmented workflows and policy-controlled decisions.

The discussion specifically mentioned adjacent areas such as:

Legal
Compliance
Finance
IT

That expansion makes sense. Many enterprise processes cut across these functions already. A supplier onboarding flow, for example, may involve procurement, legal, risk, tax, finance, and IT access controls. If each department automates only its own piece without shared orchestration, friction remains.

A broader orchestration layer could create several enterprise benefits:

A more unified employee experience
Reduced duplication of workflow tooling
Better reuse of master data and business rules
Faster coordination across departmental boundaries
More scalable AI deployment through shared governance

This also aligns with how users actually experience work. Internal stakeholders do not think in terms of software categories. They think in terms of getting something done.

If orchestration evolves into a cross-functional work layer, procurement could become one of the first enterprise domains to demonstrate what that model looks like.

What Procurement Leaders Should Do Now

For organizations evaluating orchestration or agentic procurement capabilities, the discussion points to a disciplined next step rather than a grand transformation plan.

Start with a real constraint

Identify one workflow where delays, manual effort, or fragmented systems are causing measurable pain.

Define the "why" before the "what"

Be precise about the business reason for change. Faster intake is rarely the full answer.

Demand proof in your environment

Do not rely on generic demos. Use realistic scenarios, integrations, and exception cases.

Separate AI assistance from AI action

Understand exactly where automation ends and autonomy begins.

Build governance early

If teams can create agents or workflow logic quickly, oversight must scale with experimentation.

Revisit architecture assumptions

Challenge whether every gap requires another major application purchase.

Think cross-functionally

Assess whether the orchestration capability you choose could also support legal, finance, compliance, or IT workflows over time.

Conclusion

The procurement technology conversation is shifting from digitizing steps to coordinating outcomes.

That is the deeper significance of orchestrated agents and modern procurement orchestration. They are not just new tools layered onto old problems. At their best, they create a more adaptive operating model - one that connects demand intake, enterprise data, process execution, and targeted automation in a way that suites alone often have not.

The most grounded message from the discussion is also the most useful: start where the pain is, prove value quickly, and evaluate solutions based on what they can actually execute in your environment.

For procurement professionals and technical decision-makers, that is the path away from chaotic process handoffs and toward a more reliable, AI-enabled procurement function. Not a future built on hype, but one built on orchestration that works.

Source: "EP48: Orchestrated Agents: The New Operating Model for Procurement" - DPW, YouTube, Aug 6, 2026 - https://www.youtube.com/watch?v=Hjq123Cpc2Q

How Orchestrated Agents Transform Procurement