Monash Health and Heidi have partnered to co-develop and evaluate an AI translation tool for healthcare settings, now being trialled across Monash Health's hospitals and clinics.
Real-time trial
The project centres on a real-time language translation system designed to support communication between patients, families and clinicians during consultations, inpatient care and other interactions. It will initially cover 10 commonly spoken languages.
The trial is taking place across a health network where 45% of patients speak a language other than English at home. The work is focused on clinical environments, where privacy, security, governance and safety requirements differ from those of consumer technology.
Monash Health is one of Australia's largest health services, with more than 250 services across over 50 locations and in the community in south-east Melbourne. It provides 4.5 million episodes of care each year and employs 24,000 people.
Hospital network
Heidi says its products support more than 2.7 million patient interactions each week in 110 languages across 190 countries. It has raised USD $96.6 million from investors including Point72 Private Investments, Blackbird, Headline, Latitude, Possible Ventures and Archangel.
The translation tool is being evaluated in real clinical settings rather than in a laboratory or simulation. The application is designed to meet hospital standards for privacy, security and clinical safety.
Interest in translation technology in healthcare is growing as providers seek to improve communication with culturally and linguistically diverse communities while managing workforce pressures. Hospitals have traditionally relied on in-person or phone interpreters for many conversations, particularly those that are sensitive or clinically complex.
At Monash Health, professional interpreters will continue to be used, especially for complex care or sensitive conversations. The trial is intended to test where AI translation can support day-to-day communication without replacing those services.
Clinical focus
Adjunct Professor Robin Mann outlined the health service's rationale for the project.
"At Monash Health, we know how important communication is to our patients. Communication allows our clinicians and those in our care to build trust, and ensure patients feel empowered in their own care journey," said Adjunct Professor Robin Mann, Chief Digital and Information Officer, Monash Health.
"By supporting patients and clinicians to better understand one another in the moment, the technology seeks to enhance safety, reduce misunderstandings, and improve shared decision-making," said Mann.
The collaboration also gives Heidi an opportunity to test its software in a large public health service, where deployment standards are likely to be more demanding than in many private-sector settings. For healthcare AI suppliers, evidence from clinical use has become increasingly important as hospitals and regulators scrutinise product claims more closely.
Dr Thomas Kelly said the partnership would help shape the tool through direct use in care settings.
"Heidi exists to give Australian clinicians back the capacity to focus on their patients. When a clinician and patient can truly understand each other in their own language, the quality of that care goes up. Trust builds faster, decisions get made together, and patients leave feeling genuinely heard. Partnering with Monash Health means we're building this with the evidence and rigour it deserves," said Dr Thomas Kelly, Chief Executive Officer and Co-Founder, Heidi.
Industry push
The partnership will examine how language and AI tools can be applied responsibly in patient care at scale. The question has become more pressing for hospitals as they weigh the potential benefits of automation against risks related to accuracy, bias, privacy and oversight.
"Through co-development, testing and proof in practice, Monash Health and Heidi will jointly explore how advanced language and AI-enabled technologies can be responsibly applied to support inclusive, patient-centred care at scale," added Mann.