Pasito

The differentiating capability is model work: agents read unstructured plan documents and employee census files, extract and standardise them into a common schema, and construct the plan configuration that everything downstream depends on. That is what carriers are buying and it cannot be done with rules against documents that arrive in a hundred formats.

Underneath sits an earlier and still operating layer of benefits microsites, communication campaigns and enrolment experiences, which functions without models. Apply the removal test and a working benefits experience platform remains, so this lands with the workflow vendors rather than the model native ones.

Automation is described across the whole lifecycle, covering plan construction, sales, quoting, marketing, enrolment, support and claims operations, and a conversational assistant answers employee benefits questions around the clock without a person involved. No review gate, approval step, confidence threshold or escalation path is described anywhere.

That matters more here than in most workflows because plan construction errors are systemic rather than individual: a single misconfigured plan propagates to every employee who enrols under it and is typically discovered at a claim, months later, by someone who had no way to check.

Publishing 98 percent plan construction accuracy against a 74 percent benchmark from a named industry research body is the strongest measurement practice in the insurance cluster of this index, because the comparison is anchored to an external reference a buyer can look up rather than to the vendor's own prior state. That single figure does more for a validator than pages of description elsewhere.

What surrounds it is missing: no methodology for how accuracy was assessed or on what sample, no error analysis by document type or plan complexity, no model documentation, and no support statement for a carrier's own validation.

Customers are named at the top of the market, including two major life carriers and a group benefits insurer, alongside large benefits consultancies, retirement advisers and recordkeepers, with deployments reaching thousands of employers. Growth is stated at fifty times recurring revenue over a year.

The standout is the accuracy claim: agents are reported to reach 98 percent accuracy in plan construction against a 74 percent industry benchmark attributed to a named insurance research body, which is a falsifiable figure anchored to an external reference rather than a self declared adjective, and almost nothing in this index does that. A further employer side figure reports enrolment decision support doubling health savings account elections.

Agents are described as customised for insurance and benefits and configured against each client's own plans, which scopes behaviour to the customer, and the published accuracy measurement shows performance is being tracked rather than assumed. What is missing is the data layer beneath it.

Nothing identifies which model providers process plan documents and census files, states whether one carrier's plan structures inform capabilities serving another, or describes retention once a plan has been constructed, and a stated intention to let clients build proprietary agents inside the workspace makes the boundary question more pressing rather than less.

The data here is unusually sensitive and the silence around it is the gap. Employee census files carry names, dates of birth, salary, dependants and coverage elections, and group health plan administration sits inside the federal health privacy regime, which imposes specific obligations on a service provider handling protected health information. A competitor in this index states health regime readiness explicitly. No published privacy framework, health information position, retention schedule or subprocessor list was located here.

No trust centre, enumerated certification list, attestation scope or audit period was located in this pass. That absence is more consequential here than for most vendors in this index because the platform holds employee level census and health benefit enrolment data for thousands of employers, which is precisely the material a carrier or a benefits consultant would require independent assurance over before signing, and the largest named customers would certainly have demanded it privately.

Pasito supplies software and holds no licence, which is expected, and its regulatory grounding is thinner than the domain warrants. Group benefits sit inside the federal employee benefits statute with its fiduciary duties, health coverage carries reporting obligations and health privacy rules, and retirement benefits bring their own fiduciary regime for the advisers and recordkeepers named as customers. None of those frameworks is addressed publicly, no supervisory instrument is named as a design target, and no formal admission process has been passed.

One marketed outcome deserves scrutiny rather than applause. Decision support is credited with doubling health savings account elections, and an election rate is a conversion metric, not evidence that the recommendation served the employee. A high deductible plan paired with a savings account suits a healthy employee with cash reserves and can be the wrong choice for someone managing a chronic condition or living close to their income.

Nothing public describes what the recommendation optimises for, whose interest it represents when the employer, the carrier and the employee diverge, or how recommendation quality varies across household circumstances.

Two things sit on the right side of the ledger: publishing an accuracy figure creates an expectation a customer can measure against, and employee facing recommendations are advisory, so a person can enrol against the suggestion. Neither binds the vendor.

No accuracy guarantee, no remediation term, and no described process where a misconstructed plan is discovered at a claim, which is the failure mode that matters, since by then the employee has already made a coverage decision on the strength of a configuration nobody re examined.

Agents are described as customised for insurance and benefits and configured on each client's own plan documents, which tells a buyer the behaviour is scoped to their data. Beyond that the chain is undisclosed. No model providers are named for document extraction, schema construction or the conversational assistant, no subprocessor list is published, and nothing states where plan documents and employee census files are processed, which is the question a carrier's vendor review would ask first given the personal and health information involved.

The ingestion side is the hard part and it is the product: taking plan documents and census files in whatever form a carrier, broker or employer holds them and producing one structured layer that downstream workflows can run on. Deployment inside major carriers implies real integration into their operations, and stated direction includes white labelled experiences and customer built agents inside the workspace.

What was not located is the connector detail a benefits buyer would ask for, with no named benefits administration, payroll or recordkeeping system integrations, no public developer documentation and no partner directory.

Delivery is cloud hosted software as a service serving domestic carriers, brokers and employers, so cross border complexity is limited. Residency still matters given the content, since employee census and health enrolment data are held alongside carrier plan libraries. No hosting regions, tenancy separation, residency options or subprocessor chain were located in this pass.

No rates, tiers, billing unit or minimum were located. The buyer set spans four very different economics, from a national carrier to a single employer running one enrolment a year, and nothing public indicates whether pricing follows plans constructed, employees covered, seats or enterprise agreement, which leaves the smallest of those buyers unable to size it at all.

Four audiences are addressed with separate material and different arguments: insurance carriers automating document and plan production, brokers and benefits consultants, financial institutions including retirement advisers and recordkeepers, and employers running enrolment. Product lines cover group health, life and retirement, which spans both the insurance and the retirement savings sides of the benefits market. The boundary is that market itself, with nothing addressing property and casualty, banking, lending, payments or capital markets.