Families can filter care categories, compare example caregivers, agree a plan, review changes and retain visit notes. This browser-based prototype develops the original four-category vision. Live accounts, screening, payments, messaging and staffed operations are the next implementation requirements.
Prioritise the shared care-plan and change-review workflow, then prove service readiness. The strongest differentiation hypothesis is that families and caregivers can identify the effective instructions and schedule more reliably across changes and repeat visits. Existing providers already offer matching, recurring care, messaging, screening and, in Sunhouse’s case, visit notes reporting. A defensible advantage would require measured usability gains, retained relationships, dependable local supply and permissioned longitudinal data—not a longer feature list.
Dated requests, profile details, plan acceptance, linked visits, change/cancellation review, checklists, visit notes, assigned follow-up, feedback, next-week drafts and readable reports.
Authenticated participants, server permissions, transactional writes, privacy controls, screening and commercial routing.
Real supply and notifications, category-specific operating readiness.
Original-plan references use PDF pages in the supplied Kinova AI Ltd. Business Plan (S19). Official competitor pages reviewed November 10, 2026. “Demonstrated” means local workflow, “Partial” means limited implementation, and “Planned” means absent. Public-page silence does not establish a competitor gap.
| Product feature / summary | Original business plan | Achievement in this preview | Canadian competitor comparison | Kinova differentiation | Gaps and strengthening / expansion |
|---|---|---|---|---|---|
| Request and household instructions | pp10–11, 32: request-to-close workflow; child care includes snacks, homework and supervision. | DemonstratedDated care weeks, all seven days, times, language preference and instructions; validation and local saving. All four care categories, support types and authority confirmation. | CN: needs intake and matching. CC: job posts and applications. | A specific routine is carried through the plan, visit and summary. This is a design emphasis, not demonstrated superiority. | P0 · Product: evaluate four adult infant-care arrangers and four senior-care arrangers or recipients; record completion and misunderstood task boundaries. Add structured permissions for outings, food and visit notes. |
| Caregiver comparison and matching | pp10, 30, 32: search, match, rates, caregiver scoring and supply management. | Partial500 example profiles use compact expandable results and detail pages, category/support-type/language filtering, and schedule eligibility checks; no live inventory or ranking. | NS: availability/language/rate filters. CN: nearby matches. | Explain the uncovered afternoon rather than showing an opaque match score. Matching itself is established competition. | P0 · Engineering: availability records with time zone, overlapping bookings, travel buffers and expiry. Prove zero double allocations in concurrent tests; then evaluate ranking with consented data. |
| Mutual plan acceptance | pp32: Request → Match → Confirm → Deliver → Close → Continuity. | DemonstratedFamily proposal, caregiver acceptance or return for revision; acceptance creates linked visits. Client-side participant guards. | SH: requests, caregiver acceptance and booking confirmation. | Carry exact household instructions into the accepted record. Public sources do not establish whether competitors lack this. | P0 · Engineering: immutable plan revisions, server authorization, idempotency keys and atomic acceptance. Show two real authenticated test accounts and denial of a third household. |
| Versioned schedule changes | pp10, 30: integrated workflow and continuity; explicit two-party revision control is a proposed refinement. | DemonstratedPending change leaves the current visit and estimate intact. Accept updates both; decline keeps the original. Stale displayed revisions are rejected locally. | SH: integrated messaging and repeat arrangements; detailed revision semantics not established from the public page. | Strongest testable hypothesis: fewer misunderstandings when a changed time is separate from an agreed time. | P0 · Product/engineering: compare change tasks against existing alternatives; require all participants to identify the effective time. Cancellation review is implemented locally. Add changed duration, caregiver replacement and transaction-level concurrency. |
| Visit timeline and visit notes | pp10, 30, 32: tracking, delivery, closure and continuity. | PartialAccepted times drive the visit timeline. Caregiver records category-specific completed tasks, saves a required note and completes the visit once. No actual check-in, time verification or attendance measurement. | SH: public end-of-shift-report testimonial. | Visit notes alone is not unique. Connecting unfinished tasks and permissions to the next visit may improve continuity; this remains untested. | P1 · Product: assigned follow-up and completion results now work locally. Add declined-task reasons. Test whether the next caregiver correctly identifies outstanding work without reading a full chat history. |
| Feedback and repeat requests | pp10, 30, 32–33: rebooking, quality metrics, retention and reliability. | PartialFamily feedback after a completed visit; repeat creates a next-week draft with the original requested routine and archives the prior arrangement. It does not copy a one-off changed start time. | CC: caregiver reviews and concern reporting. SH: recurring care. | A repeat draft preserves household context. Repeat booking and reviews are not exclusive capabilities. | P1 · Product/data: choose recurring baseline versus last accepted week explicitly; link repeat lineage and reasons. Measure repeat intent separately from actual repeat use and report denominators. |
| Costs, payments and payroll | pp6, 11: 15% commission; p35: 13% child-care forecast assumption. | PartialIllustrative $25 hourly compensation plus proposed 15% family fee; accepted changes recalculate totals. Readable cost CSV and printable summary. No payment processing. | CN: invoicing and payroll tools listed. CC: paid memberships. | Separate compensation and family coordination fee in the explanation. This is transparency, not evidence of lower cost. | P0 · Commercial/legal/engineering: settle operating model and fee discrepancy; map taxes, payroll, refunds and chargebacks. Reconcile a provider test transaction end to end before enabling live payment. |
| Screening and service eligibility | pp30–32: verified ID/KYC, background checks and screening cadence. | PlannedNo supplier integration, identity verification, background check, expiry enforcement or screening badge is implemented. | SH: interview, reference and CPR screening claims. NS/CN: background-check features. | No current advantage. Incumbent screening is a material operating benchmark. | P0 · Trust/safety: select providers; define check types, scope, expiry and access. Demonstrate that missing/expired required checks block an assignment, with review and correction paths. |
| Language and accessibility | pp10, 30, 32–33: bilingual/multilingual onboarding, Korean newcomer audience and accessibility. | PartialLanguage preference is stored; interface is English. Responsive layouts and labelled controls are present. No Korean translation, assistive-technology audit or WCAG compliance claim. | NS: language filtering. ND: tailored home-care service. | Korean/English discovery audience is a focused go-to-market choice, not exclusive supply or a language moat. | P0 · Content/design: professionally reviewed Korean interface and consent text; keyboard/screen-reader review. Test comprehension with both audiences before making bilingual-service claims. |
| Four care segments | pp10–11, 30: child, senior, special-needs and recovery support. | PartialAll four categories share request, matching, acceptance, changes, visits, visit notes and repeat flows. Infant and senior examples are populated; specialist operational protocols remain to be developed. | ND: companionship, disability and post-surgery services. Tuktu Vancouver: recurring home support with app booking, chat and payment management. CC: multiple care categories. | The original four-category vision is retained with a consistent care-plan workflow and category filters. Breadth alone is not a proven exclusive position. | P0/P1 · Operations/product: evaluate infant and senior examples; select the first launch segment using demand, caregiver availability and operating requirements. Add task-level permissions, training and escalation protocols for each category. |
| Messaging and notifications | pp10, 33: chat and multichannel support. | PlannedStructured proposals and status feedback operate locally. Earlier conversation mockups are visual concepts; this walkthrough has no chat transport or notification delivery. | SH and CC: caregiver messaging. NS: direct phone/message contact. | No current advantage. The opportunity is to make a message actionable without treating it as accepted care terms. | P1 · Engineering: conversation threads tied to requests; notification outbox with delivery states, retries and redaction. Test delivery failure without losing or duplicating the accepted change. |
| Incidents, geolocation and urgent support | pp32–33: L1–L4 incidents, geofenced check-ins, panic tool and urgent coverage. | PartialConcern text can be recorded in the example; it is not routed or monitored. No GPS, panic dispatch or staffed response. | CC: concern reporting. ND: publicly offered around-the-clock care. | A recorded concern is preparation, not safety coverage. No comparative safety claim is supported. | P0 · Operations: incident severity, ownership, response hours, escalation and rehearsal. Obtain consent/privacy design before location features; never release a panic control without tested staffed response. |
| Data, security and permissions | pp32–33: account integrity, nonmedical scope, compliance and insurance. | PartialClient-side participant checks, plain local storage, a version counter and event list. No authentication, server isolation, encryption-at-rest assurance or protected audit trail. | Competitor public marketing pages are insufficient to compare security architectures. | Transparent state behaviour is useful for review; it provides no security advantage. | P0 · Engineering/privacy: authenticated API, household-level access rules, encrypted managed storage, retention/deletion and incident controls. Test cross-household denial and concurrent updates before real data. |
| Quality analytics, training and supply | pp30, 32–33: scores, micro-badges, on-time/rebook metrics, supply ratios, surge and standby pools. | PlannedEvent history supports inspection of this example. No operational KPI dashboard, training system, reliability score or live supply management exists. | NS: profile qualifications. SH: screened local supply. | Potential future advantage would depend on useful longitudinal outcomes and reliable clustered supply, neither of which exists yet. | P1/P2 · Operations/data: define metric denominators and training evidence. Start with completion and change-handling data; evaluate demand/supply by neighbourhood and time before scoring or surge pricing. |
| Membership, employer channels and expansion | pp6, 11, 32–33: memberships, training upsells, B2B2C, Western Canada and national scale. | PlannedNo subscription, employer buyer workflow, paid training, multi-city operations or live service launch. | CC: membership model. CN: Canada-wide city coverage. | These are business-model options rather than present product advantages. A Korean/English Vancouver cohort may provide a focused entry point. | P2 · Commercial: compare family fee versus subscription with willingness-to-pay research. Explore employer-sponsored access only with identified buyers and a support-cost model; expand geography after local service economics. |
| Product evaluation records | pp30, 32–33: quality learning and service improvement; task-based adult evaluation is a current refinement. | DemonstratedParticipant code, care category, task outcome, duration and feedback save locally and export as readable CSV. | Public competitor marketing does not provide a comparable research-instrument specification. | Makes the infant/senior evaluation approach usable. It does not establish market validation. | P0 · Research: run the proposed adult sessions, review observed barriers and retest changes. Keep internal QA observations separate from customer findings. |
| Assigned follow-up | pp30, 32: quality and continuity; explicit owned tasks are a current refinement. | DemonstratedCompleted visit notes support practical next steps with a family or caregiver owner. Only the assigned view completes an item. Open items carry into repeat requests. | Competitor marketing supports visit notes and recurring arrangements; exact follow-up semantics require direct evaluation. | Connects a note to an owner and a result, rather than relying on free text alone. | Evaluate retrieval and completion with adults; add production notifications and household permissions. |
| Provider enquiry directory | pp10, 32: care discovery and local supply; an independent directory is a current extension. | DemonstratedEleven sourced care businesses in expandable results, category filters, saved shortlist and official enquiry links. Separate from 500 example profiles and twelve populated care situations. | Directories and care agencies already provide direct enquiry routes. Official links identify the actual business. | Adds a useful present-day research action while users explore Kinova’s care process. | Recheck sources, invite listing corrections and obtain explicit commercial participation before integrated booking. |