How Sonet defines, executes, and bills for work, and why the model is built around shared accountability.
Many AI solutions charge a flat monthly fee regardless of output. You pay for the tool whether it works for you or not. Sonet charges differently: you pay for work executed on your behalf. The more PAT does for your pharmacy, the more valuable it becomes, and the more you pay. That alignment is intentional.
For this model to work fairly for both parties, we need to agree on two things: what a Task is, and what it means for a Task to be executed.
A customer-defined workflow: the step-by-step logic describing how a specific pharmacy process should be handled. A Task is defined by the customer, approved by the customer, and scheduled or triggered by the customer. PAT does nothing until all three have occurred.
Each instance in which PAT runs a Task from start to its defined end state. An Execution is the billable unit.
The condition established by the customer's own workflow logic at which PAT stops. This may be a completed action, a hold, a queue placement, or a hand-off to the pharmacy team. All are valid end states.
An instance in which PAT fails to reach the end state due to a Sonet platform or agent error. A Failed Execution is not billable.
Because the customer defines, approves, and schedules or triggers every Task, there are no surprises. PAT does not run anything you have not authorized. You decide what it does, how often it does it, and when it starts. Billing follows execution, and execution follows your direction.
Every billing question traces back to one question: did PAT follow the approved workflow logic? The answer determines accountability clearly and for both parties.
PAT followed the logic. The logic had a gap the customer did not anticipate.
CustomerThe Task was executed. Billable.
PAT failed to follow the logic due to a platform or agent error.
SonetThe Task was not executed. Not billable.
PAT reached a step requiring clinical judgment and escalated to the pharmacy team per the approved logic.
NeitherThe hand-off is a successful execution. Billable.
When PAT escalates to a pharmacist per the approved logic, it has done exactly what it was instructed to do. Escalation is not failure. It is the workflow reaching its defined end state.
A single misaligned execution is a $0.25 question. Neither party ends a relationship over $0.25. A month of unchecked misaligned executions is a $2,500 problem, a damaged relationship, and lost time that nobody gets back.
That is the real risk. And it is why agreeing on what constitutes an execution matters as much as it does. A clear, shared definition gives both parties the language to flag issues early, stop the run, and fix the workflow logic before a small problem compounds into a billing dispute or, more importantly, a month of output the pharmacy did not want.
When this model is working correctly:
The pharmacy's workflows run reliably. Output matches expectations. When something does not work as intended, it is caught within hours, not weeks. The relationship grows because trust compounds with every Task executed correctly.
A pharmacy defines a patient intake workflow. The logic covers collecting patient information, validating required fields, and routing the record to the appropriate queue.
PAT follows the approved logic exactly. A scenario the pharmacy did not anticipate, a record missing a required field type not accounted for in the logic, is not handled as expected. All other records are routed correctly.
This is the model working as intended. A small logic gap, caught early, costs almost nothing to resolve. Left unchecked for a month, the same gap costs everyone: the pharmacy in lost operational time, Sonet in a damaged relationship. Early detection is not just a billing protection. It is how the pharmacy gets the business impact it signed up for.
This model only works if both parties are honest about what is happening and committed to fixing what is not.
That requires trust. And trust is not assumed. It is built through consistent effort, transparent communication, and a willingness to address problems directly when they surface. The structure of this relationship is designed to build that trust, because both parties have the same incentive to make it work.
Sonet grows when pharmacies trust PAT with more of their work. The path to that trust is not a sales conversation. It is execution. Every Task PAT handles correctly is evidence that the pharmacy can rely on it for something more complex. Every logic gap addressed quickly is proof that Sonet will not let problems fester. The only way Sonet makes more money is by earning more responsibility. The only way to earn more responsibility is to deserve it.
The pharmacy is never locked in. Terms are month-to-month. If PAT consistently underperforms, the pharmacy can leave. That is not a vulnerability in the model. It is a feature. It means Sonet is always working to earn the next month, not just the first agreement. The pharmacy retains accountability, and Sonet accepts that willingly because it is the right foundation for a relationship worth keeping.
Start with something simple. Let PAT prove itself. As confidence builds, add more workflows, more complexity, more volume. That is how the relationship scales: not through contract obligations, but through demonstrated performance and shared success.
per executed Task
per month minimum
(equal to 3,000 executions)
Month-to-month terms
You never pay both the minimum and the per-task rate: it is whichever is greater. Run 1,000 executions in a month and you pay $750. Run 4,000 and you pay $1,000. Pilots start with the first two weeks free.
Request a Demo →