The physician makes a request
Analyze the prescription for this 82-year-old patient with renal impairment and suggest adjustments if needed
Streamline Regulatory Compliance for Your Prescription AI
Your AI agents call Posos CE IIb-marked clinical APIs* for prescription analysis and treatment suggestion. Expand your compliant functional scope and accelerate your time to market—in Europe and internationally, powered by a 100% SNOMED-CT normalized database.
Already integrated by software vendors building the future of healthcare
.png)
CE IIb-certified APIs
CDS Hooks & MCP-compatible
All LLMs supported
Integration in 2 weeks
Our Differentiation
The only platform with CE IIb-certified APIs for personalized treatment correction and suggestion
Unlike uncertified LLM + RAG approaches, Posos provides an auditable deterministic engine carrying the CE IIb marking. Your AI agents rely on a platform that is compliant by design.
Generation
Auditability
Compliance
Hallucination risk
APIs
Deterministic (knowledge graph)
✅ Explicit rules, sources and clinical logic exposed at every step
✅ CE IIb on the full analysis and suggestion scope
✅ APIs anchor LLM reasoning and control its clinical outputs
LLM + RAG Approaches
Statistical (next-token)
⚠️ Sources cited, opaque reasoning
⚠️ Post-hoc human validation
⚠️ Present, with no safety net
Avec
Sans Posos
Données patient
Normalisées SNOMED-CT, FHIR, ICD-10/11.
Champs texte libre, formats propriétaires, silos.
Suggestions thérapeutiques
Issues d'un graphe de connaissances médicales, chaque recommandation est sourcée et traçable.
Générées par LLM, non auditables.
Conformité
API certifiées CE IIb prêts à l'emploi, conformité LAP V2 accélérée.
Roadmap interne de 36+ mois pour être certifié LAP V2 et CE IIb.
Intéropérabilité
API modernes FHIR et CDS Hooks natifs, récupération et retranscription de l’historique patient par modèles IA propriétaires.
Intégrations sur mesure coûteuses et longues, nécessitant des mappings manuels complexes et imparfaits.
Agents IA
Source d’informations médicales validées et structurée via des APIs certifiées.
Risque d’accès à du contexte non vérifié, risque d'hallucination.
Why integrate Posos ?
Why Connect Your AI Agents to the Posos APIs
Expanded Compliance Scope
Posos clinical APIs are CE IIb-marked for treatment and dosage analysis and recommendation. They cover the full range of prescription analysis use cases.
Accelerated Roadmap
No need to build and maintain your own certified prescription engine. Focus your teams on user experience and orchestration.
Auditable Results
Every suggestion produced by Posos is traceable and sourced. The physician can review the clinical logic before validating.
Native International Scalability
100% SNOMED-CT normalized medication database, structured by International Virtual Medications compatible with IDMP. Deploy your AI agents in new countries without rebuilding your clinical layer.
How It Works
How Your AI Agents Use the Posos APIs
01
02
Your AI agent determines which API calls are needed
Based on the protocol you've defined, it selects the relevant APIs: interaction checking, profile-based dosage adjustment, alternative suggestions.
03
Posos APIs execute and return structured results
Detected interactions with clinical justifications, recommended dosage adjustments based on renal function, therapeutic alternatives with dose equivalences.
04
The physician receives a summary and decides
Your agent presents the results clearly. The physician validates, adjusts, or asks for more detail. The final decision is theirs.
Our Approach
Two APIs to Cover the Full Prescription Scope
Suggest and Secure Treatments
Contextualized prescription analysis based on patient profile (age, renal/hepatic function, pregnancy, comorbidities, allergies). Drug interaction and contraindication detection. Therapeutic alternative suggestions with dose equivalences.
Secure Dosages
Personalized dosage recommendations based on patient profile. Automatic detection of overdoses, underdoses, and inappropriate dosages. Adjustments for renal and hepatic impairment.
Clinical Applications
Examples of what you can build with the Posos APIs
Voice-Assisted Prescribing
The physician dictates their prescription. Your AI agent calls the SNOMED-CT structuring API to normalize the data, then the Posos clinical APIs to verify drug relevance and secure dosages. Prescription validated in seconds.
Medication Reconciliation
Your agent automatically compares the patient's existing treatment with the new hospital prescription. Posos APIs analyze discrepancies (discontinuations, additions, dose changes) and flag interaction or contraindication risks. Structured report generated for medical review.
Perioperative Optimization
Your AI agent sends the patient's chronic treatment and surgical context to the Posos APIs. They analyze perioperative contraindications (anticoagulants, antidiabetics, etc.) and return a guideline-compliant pre/post-operative adaptation plan. The agent presents it to the physician for validation.
"Posos's expertise will enable us to prepare all future AI-enriched clinical decision support—whether LLMs or other tools—for Bow Medical. It's also a modern architecture with APIs marked as Class IIb medical devices, compatible with new AI-assisted prescribing workflows and compliant with the AI Act."
Pierre Touton
CEO — Bow Medical
Simple Integration, Regardless of Your LLM
Compatible with all models
Claude, GPT, Mistral, Gemini, LLaMA, or proprietary model. Call our APIs via REST, GraphQL, or MCP (Model Context Protocol).
Customizable protocol
You define when to call each API, in what order, based on which criteria. You retain full control of orchestration.
Sandbox available
Test your scenarios with synthetic data before committing. Full documentation and examples provided.
Up-to-date references, international reach
Medication database updated daily, SNOMED-CT normalized and structured by International Virtual Medications (IDMP). Fast, low marginal cost country replication. Regulatory monitoring handled by Posos.
Have another question?
Contact us
Request Sandbox AccessPosos carries the regulatory compliance for the full scope of prescription analysis and suggestion. Your AI agent remains a separate device, subject to your own regulatory obligations (notably under the AI Act), but it does not need to be certified as a medical device for the clinical functions it delegates to the Posos APIs.
You define a protocol (or "enriched system prompt") describing when and how to call each Posos API. Your LLM follows this protocol based on user requests. The APIs return structured results (JSON/GraphQL) that your agent reformulates for the physician.
The anonymized patient profile required for analysis: current treatments, diagnoses, biological parameters (creatinine, weight, age, pregnancy…). No personally identifiable data required. GDPR-compliant architecture with certified health data hosting.
No. Posos does not generate free text or interact with the physician. Your LLM remains the orchestrator of the user experience. Posos provides the certified clinical analysis layer your LLM calls when it needs to verify or suggest a treatment.
Between 2 and 4 weeks depending on your architecture and desired level of customization. Our technical team supports you through implementation and testing.
Posos handles regulatory monitoring and continuously updates its APIs and medication database. Changes are transparent to you. In case of a major change impacting your integration, you are notified in advance.
Yes. The Posos database is fully SNOMED-CT normalized and structured by International Virtual Medications compatible with IDMP. Semantic standardization removes language barriers and enables fast coverage replication to new countries. Posos is already deployed in France, Belgium, and Luxembourg.
Accelerate Your AI Roadmap
Sandbox available. Technical support included.
