Resolution as a Service (RaaS) is the enterprise software pricing model in which vendors charge for discrete, verifiable problems solved rather than the number of users licensed to access the software. On May 27, 2026, Veeva Systems gave that argument its clearest market validation yet, and simultaneously demonstrated precisely the gap the RaaS architecture was designed to close.
Veeva announced Falcon, an agentic platform targeting life-sciences drug development workflows. The initial use cases are specific: trial master file document intake and quality control, health authority correspondence, and safety case triage and intake. Veeva’s own framing for what Falcon does is “agentic labor.” That is not a marketing phrase. It is a commercial category declaration. Veeva is telling its customers that its software is no longer a system you operate. It is labor you deploy.
That distinction is the SaaSpocalypse in a single sentence.
Veeva Got the Direction Right. The Commercial Model Didn’t Follow.
Veeva’s Falcon announcement does not publicly define the pricing model. But if Falcon follows the broader market pattern of usage, token, credit, or inference-based AI pricing, the same commercial gap applies: activity pricing and resolution pricing are not the same model. Usage-based pricing is becoming the market’s default answer to agentic software, and it is directionally correct in one sense: it is better than seat pricing.
But usage pricing and resolution pricing are not the same model. They are not adjacent models. They describe fundamentally different commercial relationships.
Usage pricing tells a customer how much AI activity occurred. Resolution pricing tells a customer how many problems were solved.
A token is a computational trace. It is a record of inference that happened. It says nothing about whether the safety case was triaged correctly, whether the TMF document passed quality control, or whether the health authority correspondence was handled compliantly. In Veeva’s specific target workflows, those distinctions are not abstract. They are the difference between a completed regulatory obligation and a compliance exposure.
Tokens measure what the AI did. Atomic Resolutions measure what the AI achieved. The gap between those two sentences is the entire commercial argument for Resolution as a Service.
The Three Tests Veeva’s Workflows Are Built to Support
The foundational unit of the RaaS architecture is the Atomic Resolution: a discrete, verifiable unit of completed work that must satisfy three criteria simultaneously. It must be verifiable, meaning a problem was solved rather than attempted. It must be attributable, meaning the resolution is traceable to the platform’s AI execution rather than a human override or concurrent process. And it must be finite, meaning it has a clear endpoint that prevents open-ended agentic loops from consuming unlimited compute under a fixed fee.
Veeva’s Falcon use cases appear unusually well suited to Atomic Resolution pricing because they can be framed as discrete, regulated, auditable units of completed work. A trial master file document that has passed quality control is verifiable: the document either passed or it did not. The regulated context makes attribution more achievable because audit trails, validation records, and documented workflows already matter. But Falcon would still need to prove that the agent’s execution caused the result rather than merely participated in the workflow. And the completion boundary is finite: a single document reviewed, a single case triaged, a single correspondence classified and routed.
These workflows are not generic AI activity. They are strong candidates for Atomic Resolution pricing, provided the platform can prove completion, attribution, and quality at the work-unit level.
Why Regulated Industries Get There First
The conventional instinct is to treat regulated industries as late adopters: conservative procurement, slow sales cycles, complex compliance requirements. In an agentic labor market, that instinct is exactly backwards.
Regulated industries are better early markets for resolution-native pricing for three structural reasons. First, the tasks are formally defined before any vendor arrives. Standard operating procedures, validation requirements, and audit obligations already specify what a completed work unit looks like. The resolution boundary is not invented for the pricing conversation; it exists in the regulatory framework. Second, the value of correctness is material. A failed safety case triage, a compliance miss, or a regulatory filing error carries consequences that make the customer’s willingness to pay for verified completion, rather than attempted activity, structurally higher. Third, the buyer already expects evidence. The enterprise customer in a regulated vertical is accustomed to receiving audit logs, validation records, and documented workflows. Asking them to verify that an agent resolved a case rather than merely processing it is not a foreign concept. It is the norm they already operate under.
Seat-based SaaS vendors watching Veeva’s move should note what this implies for their own transition roadmap. The question is not whether your market will eventually demand resolution-native pricing. The question is whether your workflows have the definitional clarity to make Atomic Resolutions defensible at renewal. Regulated verticals have that clarity by force of compliance obligation. Unregulated verticals must build it deliberately.
What Resolution Pricing Does and Does Not Eliminate
The argument for outcome-based pricing is sometimes stated too broadly. It does not eliminate risk for the customer. It relocates and reframes one specific kind of risk: the obligation to pay for unresolved work.
Under usage pricing, a customer whose safety case triage agent processed 10,000 cases but produced incorrect classifications on 1,200 of them has consumed the tokens for all 10,000 cases. The invoice reflects activity. The liability for the incorrect classifications sits with the customer.
Under Resolution as a Service, the vendor is paid for verified completed work. An Atomic Resolution is billable only when it is verifiable, attributable, and finite. A case that was processed but not correctly triaged is not a billable resolution. The vendor’s incentive is structurally aligned with the customer’s compliance obligation, not with the volume of inference consumed.
This is the commercial architecture that Veeva’s Falcon platform points toward, even if its current pricing model has not yet arrived there. Veeva has adopted the language of agentic labor. The next layer is the economics of resolved labor.
What the Transition Requires
Moving from activity pricing to resolution pricing is not a contract amendment. It is an architectural commitment. The measurement infrastructure that makes Atomic Resolutions verifiable, attributable, and auditable has to be built before the billing conversation begins. Vendors that change the contract language without building the underlying architecture will discover at the first billing dispute that they cannot prove what they resolved.
In regulated industries, that audit infrastructure is not optional, and it is not a future requirement. FDA-regulated environments already operate under electronic-record, validation, and audit-trail requirements, including 21 CFR Part 11 where applicable. What is less settled is how those existing obligations will apply when autonomous AI agents execute multi-step compliance-sensitive work across systems. Any agent builder selling into life sciences, healthcare, oil and gas, or financial services needs to be able to answer, before formal regulatory guidance arrives, exactly what its agent did, when, and against which inputs.
The Three-Phase RaaS Transition Roadmap begins with a Revenue Audit in months one through six: identifying which workflows can be defined as Atomic Resolutions, calculating the cost-to-serve at the resolution level, and building the measurement infrastructure that makes those resolutions attributable and auditable. Vendors also need Resolution Contribution Margin discipline, which measures whether each resolution class remains profitable after the full cost-to-serve stack is included. A vendor that skips the architecture and changes only the contract language will discover at the first billing dispute that it cannot prove what it resolved.
Veeva is early evidence that the market is moving toward agentic labor. Its current pricing model is also evidence of where the market has not yet arrived. That gap will not stay open indefinitely.
The philosophical dimension of this transition reaches beyond pricing mechanics into a deeper question about accountability in autonomous systems: when AI agents perform regulated, high-stakes work, what institutional architecture makes that delegation legitimate? That argument is developed at Middle Way in AI, where the governance conditions for agentic delegation are examined outside the commercial frame.
Prescription
Map your highest-value workflows against the three Atomic Resolution criteria: verifiable, attributable, and finite. If your workflows serve a regulated vertical, your compliance infrastructure has likely already done most of that definitional work. If they do not, the definitional work is Phase 1 of the transition. The RaaS Vendor Transition Playbook walks through the Revenue Audit in operational detail, including the cost-to-serve diagnostic and the Atomic Resolution Catalog build.
The Socratic question: If your largest regulated-industry customer asked you today to produce a complete record of every agent task that was completed versus merely attempted in the last 90 days, what would your invoice say, and what would your audit trail say? If those two answers are not identical, you are pricing activity and calling it resolution.