Heidi II expands AI scribe into clinical admin work

Heidi has launched Heidi II.

The new product expands Heidi's role beyond documenting clinical visits.

Heidi II marks what the business described as its biggest product change since Scribe, its tool for documenting patient consultations. The new system is designed to handle administrative work that follows a consultation, including referrals, forms and follow-up actions that have typically fallen back to clinicians.

Heidi says its platform now supports 2.8 million patient visits a week worldwide and identifies about 10 million follow-up tasks from those consultations. Heidi II is intended to take on more of that workload across the software systems clinicians use.

Users can choose a pre-built "Routine" or describe in plain language what they need to happen. The system then plans the task and works through it across the clinician's systems, returning only the points that require medical judgement or approval.

Broader role

The launch reflects a broader push by healthcare technology suppliers to move beyond note-taking tools into software that can complete parts of administrative workflows. In Heidi's case, the product draws on deeper clinical context, connects with existing systems and can navigate software directly where a standard integration is not available.

The aim is to make the software a link between different parts of the clinical technology stack rather than a tool limited to capturing consultation notes. The system also adapts to the way individual clinicians prefer to work.

A central part of the product's design is that decisions requiring clinical judgement remain with the clinician. Heidi positioned the new release as a move beyond AI scribe software into what it called agentic AI in healthcare.

The administrative burden on clinicians has become a growing target for software developers as health services and practices try to reduce time spent on routine paperwork. Tasks such as preparing referrals, completing forms and organising follow-up steps can add to workloads after a patient consultation has ended.

By focusing on those post-visit processes, Heidi is seeking to address a broader section of the clinical workflow than documentation alone. Heidi II can both identify and progress follow-up work, while escalating the parts that need sign-off or professional judgement.

The launch comes as suppliers of healthcare AI tools face closer scrutiny over where automation stops and human oversight remains necessary. Heidi's description of the product places that boundary at the point of clinical judgement and approval.

Heidi says the platform currently surfaces around 10 million follow-up tasks each week from the 2.8 million patient visits it supports globally.

The company said Heidi II is designed to work within existing clinical environments rather than require practices to replace their current systems. Its ability to interact with different software platforms is intended to reduce the need for clinicians and administrative staff to move information between applications manually. The product can handle tasks based on instructions provided in natural language, allowing users to define workflows without creating detailed technical rules.

Heidi said this approach could support a range of administrative processes across different specialties and healthcare settings. The system is also designed to return tasks to clinicians when a decision cannot be completed without professional input.

This keeps responsibility for clinical decisions with healthcare professionals while allowing routine steps to be handled by the software. The company expects the technology to evolve as healthcare organisations identify additional processes that can be automated safely.

Heidi II also gives the company a broader platform on which to develop services around clinical workflow management and administrative automation. As adoption of AI in healthcare expands, the ability to integrate automation with existing systems is likely to remain an important consideration for providers.

Heidi's latest product therefore represents a shift from AI-assisted documentation towards software that can participate more directly in the work that follows a patient consultation.