Heidi's AI Scribe Outpaces Canva as CEO Targets Doubling Healthcare Capacity
Heidi's AI scribe strips admin so clinicians spend more time with patients, aiming to double capacity. $700m valuation, 2M sessions weekly; outpaces Canva's early growth.
Heidi's AI Scribe Aims to Double Healthcare Capacity, Outpacing Canva's Early Growth
Melbourne-based startup Heidi says it's growing faster than Canva's early years and has a clear goal: give clinicians more time with patients by stripping out admin work. The company's valuation jumped by almost $100m in its second 2025 round, now topping $700m and leading growth inside Blackbird Ventures' portfolio.
What Heidi Does
Heidi is an AI scribe for clinicians. It listens to consultations, drafts notes, and automates documentation so doctors and nurses can move through charts faster with fewer clicks.
Founder and CEO Dr Tom Kelly, who launched Heidi in 2021 while training in vascular surgery, said the focus is simple: "how can we give doctors more time to spend with their patients." He adds, "I feel the urgency to do that because I want to double healthcare capacity."
Scale That Matters to Clinics
The company reports usage in two million patient sessions per week. Over the last 18 months, it claims 73 million consultations across 116 countries in 110 languages.
Team size has almost quadrupled in the past year. Heidi also points to partnerships with major cloud providers, noting work with Google Cloud and AWS to offload infrastructure and prioritize clinician-facing features.
Why Clinicians Care
Documentation is still the time sink on most lists. If an AI scribe reliably captures history, exam, assessment, and plan in your preferred format, you get minutes back per patient and fewer late-night notes.
Dr Kelly says the intent is to "get rid of the tasks and the bureaucracy" that stand between clinicians and care. Heidi offers a free tier with limited features and a paid plan for full use.
Practical Checks Before You Try It
- Compliance and privacy: confirm data handling for your region (e.g., HIPAA, GDPR, or Australian Privacy Principles) and where audio/text is stored.
- EHR workflow: verify integration or export paths that fit your system and templates; test auto-population versus copy/paste.
- Accuracy and edit time: run a one-week pilot; measure average edit time per note and error rates by specialty.
- Consent: align with your organization's policy for patient notification/consent to audio capture.
- Network and hardware: check mic quality and room acoustics; standardize setup across rooms to keep results consistent.
- Security: confirm encryption in transit/at rest and role-based access; review audit logs and data retention settings.
- ROI: track notes per hour, after-hours charting, and patient throughput; compare with baseline prior to rollout.
How Heidi Says It Builds Faster
According to Dr Kelly, recent advances from vendors like Google and Amazon made it easier to deliver features clinicians care about. "They don't really mind how it's built, they just want it to work and save them time."
Bottom Line for Healthcare Teams
Heidi's pitch is direct: fewer keystrokes, more patient time, and a path to higher capacity. If the reported scale and accuracy hold in your environment, the upside is clear in minutes saved per note and reduced after-hours work.
If you're planning a pilot, keep it short, measure everything, and decide fast. For teams building internal AI skills to support safe adoption, see curated healthcare-focused learning paths at Complete AI Training.
