Eolas Medical
Eolas Medical brings hospital policies, clinical guidelines and local knowledge into a shared reference platform. Ask Eolas retrieves relevant passages and generates an answer with links to the institution's source material. Clinical owners control content approval, review dates and access, while analytics reveal common questions and gaps in available guidance.
The platform serves healthcare professionals through web and mobile access, with offline access to reference content. It can read policies from systems such as SharePoint and PolicyStat, and a named hospital deployment opens Eolas from Epic. The service excludes patient information. Its strongest controlled evidence concerns simulated antimicrobial prescribing, while broader deployments cover nursing, emergency medicine and other departments.
Capability Axes
An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read
Ask Eolas uses language models to retrieve institutional guidance and synthesize answers. The wider platform also manages documents, approvals, access and offline reference material, so those functions retain value independently of the generated response.
Clinical owners approve source content, manage access and review expiration dates. Users can inspect cited sources and must verify AI responses. The workflow supports human review, but automatic confidence thresholds and escalation triggers are unspecified.
Ask Eolas combines retrieval of institutional guidelines with a pretrained language model for search and synthesis. OpenAI and AWS are named AI providers. Specific production model versions, training adaptations and update practices remain unspecified.
The privacy policy names OpenAI and AWS as AI service providers and permits changes to other providers. Inputs and outputs can be shared with them. A complete provider inventory with model retention and training terms is unavailable.
A March 2026 npj Antimicrobials and Resistance study randomized 45 clinicians across three groups. The Ask Eolas group had no prescribing errors among 15 participants, versus six and eight in comparator groups. This was a simulation at one institution.
Patient information is prohibited, and ordinary account data is retained through the account period, subject to legal exceptions. AI inputs reach external providers. Model training use and provider retention limits remain unspecified; security incident procedures do not describe AI safety response.
The application explicitly excludes patient data and is not designed for HIPAA patient processing. A service privacy policy covers user accounts, inputs and AI providers. Institutions use the platform for professional knowledge rather than patient records.
Independent penetration testing occurs at least annually. The security program follows SOC 2 criteria, which is different from holding a SOC 2 attestation. The public trust center shows monitored controls without an accessible audit report.
The service is restricted to qualified healthcare professionals and reference use. AI responses are expressly excluded from medical advice, diagnosis or treatment recommendations, leaving clinical judgment with the user rather than an autonomous device.
The antimicrobial evaluation uses staged case complexity, a second reader workflow and retained query records to examine decision support. It provides a structured evaluation process, while comparative demographic fairness and separate model behavior audits remain unestablished.
The prescribing simulation provides a measurable error result with a stated comparator and denominator. Citation links and audit history support review. These mechanisms do not create a contractual obligation to compensate for or remedy incorrect clinical output.
Western Health and Social Care Trust opens Eolas from Epic during its Encompass rollout. PolicyStat and SharePoint supply reference content. This supports access to guidance, but does not establish reading or writing patient records through an EHR interface.
The service uses AWS databases in Ireland, with regional arrangements available by health system. Mobile reference access can work offline. Tenant isolation and the complete cross border route for external model processing remain unspecified.
Teams can start with a free plan for up to 20 users. Paid options include £1 per user per month and £2,500 annually for up to 250 members. Their overlap and AI inclusion are unclear; health systems receive custom quotes.
Named deployments span emergency medicine, pharmacy, maternity, pediatrics and intensive care. Ask Eolas has controlled evidence for antimicrobial prescribing simulations. That result does not establish equivalent AI performance across the full clinical and nursing offering.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | BAA Tier | Implementation | Source |
|---|---|---|---|---|
|
£1 per user per month; AI inclusion unconfirmed
|
Team size and annual subscription; health systems by custom quote. | Patient data is prohibited; the service is not designed for HIPAA patient processing. | Content migration is a listed feature; no separate fee is stated. | Vendor Published |
A free plan covers up to 20 users. A separate £2,500 annual team plan covers up to 250 members. The overlapping plans do not establish a single AI subscription price.