Investor room
Vision package
Enter through workforce utility. Connect the clinician’s career network to an agent-operated facility system. Build operating intelligence from permitted, accepted work.
See the next milestones01 · The company model
Rōvn’s destination is a clinician career network connected to an agent-operated facility operating system, with operating intelligence developing underneath both.
Start with useful workforce work: an open role, a qualified person and an actual first worked shift. The clinician’s identity, reusable evidence and opportunities belong in that starting experience. The facility gains a connected process across its departments while each department retains its authority.
Begin with sourcing, matching, interviews, offers, full verification, onboarding, readiness and first shift. Extend into continuing workforce operations, then other connected practice workflows.
A career home, reusable record and bounded copilot from the start. Relationships, referrals, communities and knowledge-sharing can grow around useful professional activity.
Connect the situation, requirements, decisions, actions, exceptions and outcomes. Use permitted history to understand current work and, with evidence, anticipate what may prevent the next outcome.
A proposed cycle to test. Source checks, facility-confirmed facts and self-reported information retain distinct labels. Private facility information stays within that facility’s permitted context.
Follow the operating loop →02 · The career network
The larger ambition is a healthcare professional network built on verified work. The reason to join starts with utility: pursue an opportunity, carry useful evidence forward and keep a career record current.
The clinician’s agent serves the clinician’s goals and consent. Facility agents serve their own authorized mandates. A shared workflow connects them without merging those interests or transferring decision rights.
The community can grow around that activity: colleagues, referrals, professional groups, messaging and knowledge-sharing. Reputation would use attributable work history and references, with evidence status visible and appropriate moderation. A clinician is never reduced to a hidden or permanent score.
An early test: a nurse authorizes a named second facility to see an eligible check. That facility reviews its source and freshness, applies its own requirements and records whether it accepts the check. Reuse must save measurable work.
The next tests are repeat use, maintained records, useful relationships and relevant opportunities after the first start. Community development can progress alongside facility workflows as participation and moderation support it.
Career utility belongs in the initial product plan. The wider community is planned; this room does not claim an active professional community today.
03 · Operating intelligence
The first question is practical: what could prevent this facility from reaching its agreed outcome, and what authorized work would help?
Illustrative future use: a facility expects three starts next week. The proposed system compares qualified availability, current requirements, resource constraints and unresolved dependencies. It shows uncertainty and feasible options for the authorized team; it does not invent an eligible worker or make the start decision.
Operating a workflow does not grant rights to repurpose its data. Career-sharing consent is not training permission. Professional data is not for sale.
Each proposed intelligence product needs its own permitted purpose, data rights, privacy review, buyer and operating evidence. The current planned business model remains a facility subscription per certified bed.
Inspect the proposed record and architecture →How the facility becomes understandable
The first phase is not an attempt to automate every department. It is a disciplined way to understand how a facility reaches outcomes, activate one authorized workforce workflow and leave an operating asset the facility can inspect.
| Output | Why it matters to the next workflow | What acceptance looks like |
|---|---|---|
| Operating map | Shows departments, decision rights and cross-department dependencies around a real outcome. | Affected owners confirm their portion; gaps and disputes remain visible. |
| Systems and evidence map | Prevents a source, identifier or destination from being assumed authoritative. | Source and destination owners confirm responsibility; sampled identities and totals reconcile. |
| Readiness picture | Turns a cohort into explicit accepted evidence, unresolved exceptions and named owners. | Incomplete, withdrawn and delayed cases remain counted instead of disappearing from the story. |
| Effort and economic baseline | Provides a denominator before claiming time or cash value. | Operations and finance accept the period, endpoints and definitions; unknown inputs stay unknown. |
| First operating outcome | Tests whether a bounded workflow reaches an accepted destination under real authority. | Named decision owner, permitted action, action receipt, recovery exercise and completion evidence. |
| Expansion queue | Turns broader facility knowledge into a sequenced list rather than an unbounded roadmap. | Each candidate names an outcome, owner, dependencies, rights and evidence required to activate it. |
That mapping can extend across the facility while the workforce path proceeds. Missing detail in an unrelated future department should not delay an otherwise eligible hiring workflow. Existing systems retain their record responsibility until a replacement is separately validated; observing a workflow is not permission to collect every record it touches.
This is a proposed operating-knowledge workstream. It does not amend a signed pilot’s scope, dates, cohort, pricing or release eligibility.
A rule for expansion
A broad destination becomes credible only when each expansion can explain what outcome it owns, whose authority it serves, what source is authoritative and how the organization will know the package helped rather than created more coordination work.
Patient, clinical and marketplace packages retain their specific written expansion decision, data/security review, demonstrated demand and release evidence. Billing coordination, time/pay interfaces or clinical task preparation do not make Rōvn a payer, employer or clinical decision maker.
04 · One workplace
The destination spans the practice. Delivery advances through bounded operating packages with an accountable owner, clear permissions and an accepted outcome.
| Work to coordinate | What comes next | Evidence to earn |
|---|---|---|
| Workforce entry | Open role through full verification, readiness and a first worked shift, with clinician career utility alongside it. | An owner-accepted operating map, reconciled starting state, accepted handoffs and a timed outcome. |
| Continuing workforce | Renewals, coverage, float workflows, schedule preparation, staff communication and time/pay interfaces. | Permitted integrations, acknowledged changes, exercised recovery and useful repeat use. |
| Labor and capacity planning | Demand, qualified supply, resource constraints and labor economics. | Prospective forecast quality, feasible proposals, accepted decisions and net economic value. |
| Whole-practice packages | Patient access and referrals, appointments and resources, clinical task support, revenue-cycle coordination, procurement, inventory, maintenance and administration. | Demonstrated demand, a named owner, defined data routes, the required written expansion decision and data/security review, and release evidence for each package. |
Career relationships and community can develop alongside these facility stages. They do not wait for the whole-practice destination, and their adoption and moderation need separate evidence.
A longer-term specialist ecosystem could contribute through versioned interfaces, scoped context, accepted handoffs and explicit support and recovery responsibilities. This is planned architecture, not an announced partnership or marketplace.
Existing source systems keep their record responsibilities until a replacement is separately validated. Mapping a department is not permission to ingest its patient records or activate its workflows.
The launch target and any contracted pilot scope do not expand with this vision. Current product demonstrations and the dated execution plan remain separate.
05 · Earning it
The first proof is an end-to-end start at a facility: a signed scope, a measured baseline, a complete operating record and a result the facility can inspect.
Smaller facilities are part of the intended delivery design. A clear first outcome, modest onboarding effort and sustainable implementation and support costs must be demonstrated, not assumed.
Company record and founder-set goals above: October 1, 2026. Letters express interest; they are not customer contracts or revenue. Delivery elaboration: October 7 planning direction.
06 · Our commitments
07 · The far horizon
These are long-term visions. Each would require new capabilities, institutional acceptance and its own evidence before it could become a service.
Our 2030 goal is a same-day start wherever the receiving facility accepts the relevant record. The beneficiary is a clinician who has already done the work once and a facility facing a real, time-bound opening.
The mechanism is not a universal clearance: it is a clinician-authorized package of eligible evidence, its source and freshness, matched against the receiving facility’s own requirements and a named start decision. Required checks still apply; the receiving facility may recheck, decline or require something new. This depends on source terms, interoperable evidence, clear consent, facility acceptance policy and the operational ability to confirm a real first shift.
The next falsifiable proof is modest: one named second facility records that it accepted one eligible check, with the clinician’s permission and the time and repeated work measured.
The beneficiary is a clinician whose career currently fragments into repeated uploads, expiring requirements and disconnected job processes. Our vision is a free agent that helps maintain a clinician-controlled record, surface renewal work and explore relevant opportunities throughout a career.
Its mechanism is a bounded mandate: the clinician sets purpose, recipients and limits; sees what was prepared or sent; corrects information; exports the record; and can revoke future sharing. It cannot make an employer’s hiring decision, send outside the agreed mandate or convert a career record into a hidden rank. A useful career agent depends on source-backed evidence, reliable reminders, honest opportunity provenance, consent controls and a recovery path when a record changes.
The next evidence is repeated voluntary use: clinicians maintain their record, complete a consented task and report that it reduced repeated work without creating new friction.
The beneficiary is the operator who needs capable tools without losing the answer to “who was allowed to do this?” Our far-horizon vision extends the record of identity, scope, evidence, mandate, action receipt and responsible human to software agents and, eventually, robots doing healthcare work.
The mechanism is a task-specific authority contract: permitted actions, recipient or destination, time limits, stop conditions, escalation and a named accountable person. A more capable model does not become a decision maker by default. This direction depends on reliable identity, constrained interfaces, evaluation, auditability, institutional policy and the clinical and regulatory requirements applicable to the task. It does not authorize autonomous hiring, credentialing, privileging, scheduling or clinical judgment.
The next falsifiable proof is narrower: a permitted administrative action class is exercised in a controlled setting, with approval, destination receipt, exception recovery and human takeover all visible.
The beneficiary is a facility leader deciding whether a future staffing or capacity shortfall needs action before it becomes an operational failure. Our vision is a forward view of demand, qualified supply, dependencies and resource constraints that prepares feasible options for the authorized team.
Its mechanism links a forecast to a declared horizon, inputs available at prediction time, a simple operating baseline, uncertainty and the specific decision it could support. A forecast is not permission to assign people, weaken a requirement or claim a causal effect from a correlation. It depends on authorized, time-indexed operational history, calibrated evaluation, policy-aware feasibility checks and a manual fallback when the result drifts or is not useful. Facility context remains inside that facility; private data, policies and pipelines do not train cross-facility weights or aggregates.
The first research test remains a forecast from public staffing data, evaluated against a simple baseline with its error shown.
The beneficiary is a clinician whose legitimate career evidence has to be rebuilt each time an employer, state or jurisdiction changes. Our vision is useful evidence recognized across states and, over time, countries, while the receiving authority keeps its legal and professional judgment.
The mechanism is portable, source-backed evidence with provenance, correction and selective disclosure—not a claim that Rōvn can grant licensure or compel recognition. It depends on the receiving employer, licensing authority and jurisdiction; on formats that preserve source and freshness; and on rules for language, identity, scope of practice and local requirements. One facility’s decision never transfers to another, and no clinician receives a universal score.
The immediate proof remains a single permitted handoff between two facilities. Expansion across jurisdictions would require distinct acceptance evidence and institutional agreements before it could become a service.
Our far-horizon vision is a permissioned operating record that increasingly capable AI, including future superintelligent systems, could use to perform healthcare workforce work with a named person making every consequential decision.