servei d'agent de transcripció de veu per a la consulta d'atenció primària als Centres de Salut Integral de Referència (CSIR)

The contract has a duration of 12 months. The foreseen end date is October 31, 2027. The start date will be stipulated in the contract.

The contract is divided into the following phases:

• Phase 2: Deployment to CSIRS (months 2-4)
• Phase 3: Operation, monitoring, and closure (months 4-12)

The contract can be extended for an additional 12 months. The contract's end date will not change even if its formalization is affected by a special appeal in contracting matters.

The core scope of work for this tender is the provision of an Artificial Intelligence (AI)-based voice transcription agent service, delivered in a Software as a Service (SaaS) model. This service is intended to support primary care consultations across the 27 Integral Reference Health Centers (CSIR) of the Catalan Institute of Health. The scope encompasses the implementation, integration, functional and technical support, training, maintenance, updates, and ongoing evolution of the solution for the entire duration of the contract. A crucial objective is to ensure that, upon contract completion, the ICS possesses all necessary knowledge, technical documentation, and operational data for the seamless continuity or substitution of the service, safeguarding against any loss of clinical information.

The Statement of Work summary and core deliverables include:

• Project Phases: The project is structured into three phases:
• Phase 1: Initial Planning and Design (Month 1): This involves workshops with clinical and technical stakeholders to define expectations, risks, and success criteria. It also includes establishing coordination and communication protocols, setting key performance indicators (KPIs) for implementation, defining incident and risk management procedures, and designing the tool's configuration for each professional category within the CSIRs. Deliverables for this phase include a project plan, an indicator catalog, professional category templates, and an incident management procedure.
• Phase 2: Deployment to CSIRs (Months 2-4): This phase entails a staggered deployment across all 27 CSIRs, with specific plans for each center, aiming for full deployment by the end of month 4. It also focuses on a smooth transition plan to minimize disruption to care activities and consolidating the maintenance and support model for solution sustainability.
• Phase 3: Operation, Monitoring, and Closure (Months 4-12).
• Key Deliverables and Requirements:
• Integration capabilities with the Ambulatory Health History (HES) to allow for validated report dumping, structured field association, and automated completion of clinical catalogs (e.g., problems, medications, vital signs).
• A comprehensive training plan for all professionals at the 27 CSIRs, featuring accessible materials, both in-person and virtual sessions, and specialized technical training for center referents.
• A functional and technical support service provided throughout the contract's term, capable of assisting multiple entities (all CSIRs).
• Collection of usage indicators and integration of this data into the ICS Datawarehouse.
• Storage of conversations and transcriptions in a central repository, subject to express authorization from ICS and adherence to current data protection regulations.
• Provision of complete technical documentation and a service return plan at the contract's conclusion.
• Mandatory language support for Catalan and Spanish, with the capability to support additional high-prevalence languages such as English and French, which will be valued.
• A final consolidated report, incorporating collected evidence, suitable for the evaluation of the CSIR model by relevant health authorities including the Department of Health, CatSalut, AQUAS, and CAIROS.
• The solution must offer multi-platform adaptability (PC, tablet, mobile) with a responsive design, feature a modular and extensible architecture to accommodate new functionalities, and demonstrate scalability for high data volumes with real-time transcription latency under 1 second.