Built to be trusted where care actually happens.
Expanding access to the next billion people only works if the system earns trust. Here is how we think about responsible AI, privacy, patient self-service, standards, and self-sufficiency — at a mission level, in plain terms — followed by answers to the questions buyers and care teams ask most.
Responsible AI
AI drafts and organizes. It never signs.
We use AI with graduated autonomy. It prepares, prioritizes, and organizes the work that flows into a clinic, but it does not act on its own. A clinician reviews, decides, and signs — and remains accountable for every clinical decision.
- AI drafts and organizes; it does not diagnose, treat, or sign.
- A clinician is accountable for every clinical decision and signature.
- Autonomy is graduated — the human stays in control of what matters.
Privacy by design
Personal health information is handled with care.
Privacy is a design constraint, not an afterthought. The platform is designed so personal health information can be de-identified before any external AI processing, and a local-first option lets a clinic keep its data on-site.
- Designed so personal health information can be de-identified before any external AI processing.
- A local-first option means data can stay on-site, under the clinic’s control.
- Access is scoped by role, with audit trails over who did what.
Patient self-service
Patient access is connected to clinical governance.
Patients can use the portal to share records, request care, message the clinic, review released results and visits, and request refills. Those actions are part of the same governed workflow: patient-submitted information remains reviewable, clinical decisions remain clinician-owned, and activity stays auditable.
- Patient actions stay connected to the clinic record instead of becoming a side channel.
- Patient-submitted records are reviewed before they become trusted clinical information.
- Refills, results, messages, and care requests preserve care-team ownership and auditability.
Standards alignment
Designed with reference to recognized standards.
The platform is designed with reference to a wider standards ecosystem: global public-health guidance, national ministry protocols, specialty-society recommendations, interoperability standards, terminology systems, and programme reporting requirements. This is alignment, not endorsement: citing a standard indicates the guidance the platform is designed to support work against — it does not imply review, endorsement, or affiliation by those bodies.
- Designed with reference to recognized global, national, specialty-society, and clinical standards.
- Implementation is expected to adapt to national protocols, essential medicines lists, and reporting rules.
- Alignment is a design reference, not a certification or an endorsement.
Self-sufficiency
Care should not depend on the cloud.
The platform is designed to run on a single computer, online or offline, so a clinic can keep working when connectivity is unreliable or absent. The cloud is useful when it helps — it should not be the thing care depends on.
- Designed to run on a single computer, online or offline.
- Care continues without reliable connectivity.
- The cloud is an option, not a dependency.
FAQ
Frequently asked questions
Honest, mission-level answers to the questions buyers and care teams ask most.
- Does it work offline?
- Yes. InBasket AI is designed to run on a single computer, online or offline, so care continues when connectivity is unreliable or absent. The cloud is useful when it helps — it is an option, not something care depends on.
- Who owns the data, and where does it live?
- Your clinic owns its data. A local-first option lets the record stay on-site, under your control. The platform is designed so personal health information can be de-identified before any external AI processing, with role-based access and audit trails over who did what.
- What care pathways, specialties, and programmes does it cover?
- One patient spine supports authored care pathways, programme workflow overlays such as tuberculosis, HIV and ART, malaria, immunization, and family planning, and additional specialty service lines. The same architecture enables governed pathway packs, order sets, referral loops, reporting lenses, and AI assistance without fragmenting the record.
- What can patients do in the portal?
- Portal-enabled patients can share records, request appointments or programme care, send non-urgent messages, review released results and visit summaries, and request medicine refills. Those actions stay connected to the governed clinic record; submitted records and clinical decisions remain subject to care-team review.
- Is it a medical device?
- No. InBasket AI is a clinical-workflow orchestration platform focused on expanding access to the next billion people. It is not a medical device, is not a substitute for clinical judgement, and does not provide a diagnosis or treatment. Clinicians remain accountable for every clinical decision and signature.
- How is AI used — does it make clinical decisions?
- No. AI drafts, prioritizes, and organizes the work that flows into a clinic; it never signs and it does not make clinical decisions. Autonomy is graduated, with a clinician reviewing, deciding, and signing — and remaining accountable for every clinical decision.
- Does it follow recognized clinical standards?
- The platform is designed with reference to recognized global, national, specialty-society, and clinical standards for the care areas it supports. In practice that means WHO and UNICEF guidance where appropriate, but also national ministry protocols, CDC/ACIP, USPSTF, ASCCP, KDIGO, IDF, specialty-society standards, interoperability terminology, and programme reporting rules where locally adopted. This is alignment, not endorsement.
- Who is it for?
- Frontline care teams in high-growth and under-digitized settings — clinics, hospitals, community-health networks, NGOs, and public-sector programmes — that need a practical digital workspace without large IT departments or always-on connectivity.
- How is it deployed?
- It is built to stand up on a single computer, on-site, so a team can start without a data center or a dedicated IT department. It can use the cloud when that helps, but it is designed to keep working when networks do not.
Still have questions?
Get in touch, or start a guided pilot conversation about your care setting, deployment model, and first workflows.