From Synthetic Data to the Patient's Bedside: Aindo and Mia-Care Bring the First Certified Clinical LLMs to European Hospitals

A strategic alliance to combine the speed of artificial intelligence research with the rigor of clinical medicine

In the era of generative artificial intelligence (genAI), digital healthcare is at a crossroads: moving fast to innovate or slowing down to ensure maximum safety and regulatory compliance. The collaboration between Mia-Care and Aindo demonstrates that these two souls can coexist. By combining Aindo's excellence in the generation of synthetic data with the advanced, automated regulatory governance of Mia-Care's P4SaMD platform, the two companies have paved the way for a new generation of AI-based medical software, capable of overcoming traditional obstacles of fragmentation and privacy while also accelerating access to care.

Aindo's Mission: Unlocking the Potential of Sensitive Health Data Through Generative AI

Aindo has a clear mission: to improve people's lives through responsible, safe, and reliable digital healthcare innovation. The strategic target is to transform private health data from an inaccessible limitation into a fundamental resource, safely shareable to generate evidence without undermining trust, privacy, or compliance.

To achieve this goal, Aindo has developed a patented technology to create synthetic datasets by using generative AI. This data does not contain any real personal identifiers, ensuring native compliance with the GDPR and EuroPrivacy™ certification, while preserving the statistical properties and clinical patterns of the original data.

Context and Initial Challenges: The Era of Data-Centric AI and the Hurdle of the Double European Regulatory Framework

The evolution of AI models has reached a paradigm shift dictated by Data-Centric AI. The quality of the results no longer depends solely on the sophistication or capabilities of the algorithm, but on the quality and quantity of the data on which it is trained. Yet, over 95% of the world's biomedical data is private and confined within inaccessible silos due to privacy reasons, protected by strict international regulations.

When Aindo decided to evolve its offering toward modular Clinical Decision Support Systems (CDSS), they clashed with enormous complexity: developing and managing the compliance of advanced AI-based components within a Software as a Medical Device (SaMD). Indeed, developing SaMD means having to navigate simultaneously two of the strictest regulatory frameworks in the world: the Medical Device Regulation (MDR) and the emerging EU AI Act. For the engineering team, the cognitive and documentation burden associated with compliance risked slowing down and compromising tight development and commercial launch timelines.

The Goal: A Certified High-Risk AI Medical Software

The primary goal of the project was clear and ambitious: to implement a rigorous governance framework to prepare for the certification of an AI-based SaMD that would be fully compliant with European regulations, especially the nascent EU AI Act. Specifically, Aindo aimed to transform one of its high-risk AI products into a certified medical device by 2028.

To achieve this goal without losing the agility that distinguishes an innovative company, Aindo was looking for a technological partner with strong regulatory experience, capable of:

• Absorbing the complexity of regulatory processes.

• Minimizing development risks.

• Eliminating the need to embed dedicated compliance experts within the software engineering team.

• Slashing time-to-market and time-to-compliance.

Moreover, the software architecture had to provide appropriate tools to ensure mathematically faithful data while being free of real personal identifiers.

The Solution: Mia-Care P4SaMD as Software Lifecycle Orchestrator and Regulatory Guardrail

To overcome these challenges, Aindo began a cooperation with Mia-Care, integrating the P4SaMD platform into its processes. P4SaMD is Mia-Care's proprietary infrastructure that manages the regulated software lifecycle. It’s a technological and governance enabler, serving as a holistic orchestration layer that provides:

• Golden paths: Instead of imposing rigid processes, the platform provides developers with best practices and predetermined guided pathways.

• Compliance by design: By continuously monitoring the software development lifecycle (SDLC), the platform integrates compliance mechanisms and regulatory guardrails into an end-to-end framework tailored for medical devices, with real-time alerts, scores, and recommendations.

This allows Aindo's engineering team to develop compliant software from the very first line of code, drastically reducing the bureaucratic-administrative burden and ensuring business continuity.

Why Mia-Care?

Aindo's choice fell on Mia-Care because of three fundamental pillars:

• Validated technological excellence: P4SaMD is engineered according to the highest industry standards and validated following GAMP 5 principles, guaranteeing Aindo a development environment natively aligned with the quality requirements of global health authorities.

• Integration and operational flexibility: The platform adapts to the team's operational culture, easily embedding into daily workflows (such as Git or Jira) so that compliance becomes a natural extension of the developer's work rather than an external obstacle.

• Adaptive regulatory reliability: In a constantly changing regulatory landscape, Mia-Care continuously updates the platform to align it with the latest developments in EU MDR, EU AI Act, and FDA guidelines, relieving Aindo of the burden of manual regulatory tracking.

A Blueprint to Transform Clinical Data into Ward-Ready Software

The synergy between Aindo and Mia-Care's technologies laid the foundations for a blueprint designed to securely transform real clinical data into certified software:

Layer	Technological Component	Role in the Value Flow
1. Data Layer	Real Clinical Data	Selection and curation of real data cohorts from partner institutes of excellence, such as the Centro Cardiologico Monzino IRCCS.
2. Privacy Layer	Aindo Synthetic Data	Aindo's AI layer steps in to convert real data into clinical-grade synthetic data, mathematically faithful but devoid of personal information (EuroPrivacyTM/® certified, and evaluated for privacy and utility by NIST).
3. SaMD Platform	Mia-Care's P4SaMD	The Mia-Care platform provides the AI-powered orchestration layer for the entire SDLC, supporting end-to-end design, deployment, and traceability. At the same time, it ensures full compliance with the EU AI Act and regulatory standards IEC 62304 and ISO 13485, as well as GAMP 5 guidelines.
4. Clinical Output	Certified LLM-based CDSS	The final product: a vertical Clinical Decision Support System (such as a cardiology clinical LLM) ready to be distributed as a certified medical device.

Adoption and Process: An Agile Onboarding to Turn Compliance into a Daily Routine

The adoption process was rapid and non-invasive, following an Agile methodology. It began with a one-month PaaS trial, launched on January 27, 2026, which combined the use of Mia-Care's Internal Developer Platform and specific P4SaMD features.

During this period, Mia-Care delivered targeted direct training to key figures of Aindo's technical team, introducing them to the use of the prototype and tools. In the initial POC phase, P4SaMD facilitated the creation of a complete backend service in Python natively integrated with Aindo's AI service APIs, allowing the team to quickly familiarize themselves with workflow automation and regulatory guardrails.

Business and Technological Impact: From Breaking Down Data Silos to Computational and Regulatory Efficiency

The impact of the solution is measured directly on business scalability and the speed of Aindo's scientific research:

• Documentation and traceability automation: P4SaMD produces in real time the technical and traceability documentation (from requirements to deployment) required for regulatory audits and certification processes, with an estimated 90% reduction in documentation time.

• Reduction of time-to-compliance: Thanks to automated and multi-framework compliance, the time required for new AI feature releases is reduced by 50%, accelerating innovation without introducing regulatory friction.

• Complete AI Governance: With P4SaMD, Aindo can autonomously manage and govern the compliance of all AI components (models, algorithms, etc.) of the medical software under development.

• Single source of truth: P4SaMD centralizes requirements, controls, and documentation for MDR, EU AI Act, and FDA GMLP into a single system, eliminating fragmentation and aligning all stakeholders on a shared framework.

• Personnel optimization: Aindo started the SaMD compliance journey without the need to hire additional personnel or dedicated regulatory consultants within the team.

Looking to the Future: The First Cardiology LLM in Europe and Scalability for Oncology Pathways

The first major milestone of this collaboration is already mapped out: the Cardio Clinical LLM, developed by Aindo in collaboration with the Centro Cardiologico Monzino, is on track to become the first vertical clinical language model certified as a SaMD in Europe on the Mia-Care platform.

This success represents not a final destination, but a reproducible template. As a matter of fact, the company is evolving its offering toward other modular and certified AI-based CDSS designed to support multiple clinical pathways. To begin with, the Cardio Clinical LLM will be reused as a composable asset to develop a vertical LLM dedicated to oncology and, subsequently, to map multiple other complex clinical pathways.