Harrison.ai is a Sydney-based health-tech company that builds “clinician-led” AI tools to scale global healthcare capacity. Unlike many AI startups that sell software directly to hospitals, Harrison.ai often uses a Joint Venture (JV) model: they partner with major medical groups (like I-MED Radiology or Sonic Healthcare) to co-develop the AI on massive proprietary datasets, and then commercialize the solution globally.
Their philosophy is “AI as a medical device” — building regulated, high-performance tools that act as a “second pair of eyes” for doctors, rather than replacing them.
Core Technology & Divisions
Harrison.ai operates through specialized divisions, each targeting a specific medical domain:
- Radiology (Harrison.ai Radiology / formerly Annalise.ai): Recently rebranded (Oct 2025) to align under the master brand, this is their most mature product.
- Capabilities: Their “Comprehensive Care” AI can detect over 124 findings on Chest X-rays and 130 findings on non-contrast CT Brain scans.
- Workflow: It automatically triages worklists, flagging urgent cases (like a brain bleed) so radiologists review them first.
- Foundation Model: They recently launched Harrison.rad.1, a radiology-specific multimodal Large Language Model (LLM) designed to interpret images and generate reports (currently in research/validation phase).
- Pathology (Franklin.ai): A Joint Venture with Sonic Healthcare.
- Focus: Digitizing the workflow of pathologists who traditionally examine tissue samples on glass slides.
- Tech: Uses AI to identify cancerous cells in digital pathology slides, assisting in the diagnosis of complex cancers (e.g., prostate, breast) with higher speed and consistency.
- IVF & Fertility (Ivy / iDAScore): Developed with Virtus Health and often distributed via Vitrolife.
- The Problem: In IVF, embryologists traditionally select embryos by manually looking at them under a microscope, which is subjective.
- The AI Solution: They trained an AI (“Ivy”) on thousands of time-lapse videos of developing embryos. The AI analyzes growth patterns invisible to the human eye to predict which embryo has the highest probability of resulting in a fetal heart beat.
Regulatory & Clinical Status
Harrison.ai has achieved significant regulatory success, validating their “AI as a Medical Device” strategy:
- FDA (USA): Secured 12+ FDA clearances. Notably, their solution for obstructive hydrocephalus received Breakthrough Device Designation and is eligible for NTAP reimbursement (one of only two radiology AI devices to achieve this).
- TGA (Australia) & CE Mark (Europe): Approved for clinical use in over 40 countries.
Company Profile
Founders: Dr. Aengus Tran (Medical Doctor & AI Engineer) and Dimitry Tran (Business/Strategy).
Headquarters: Sydney, Australia.
Funding: Raised approx. $112M USD ($169M AUD) in Series C funding (Feb 2025), bringing total funding to over $240M AUD.
Key Partners: I-MED Radiology Network, Sonic Healthcare, Virtus Health.
Key Use Cases
- Use Case: Emergency Triage
Description: In busy ERs, the AI reviews CT scans immediately. If it detects a stroke or bleed, it alerts the radiologist instantly, potentially saving hours of wait time.
- Use Case: Embryo Selection
Description: Instead of guessing which embryo is “best,” IVF clinics use the AI score to transfer the most viable embryo first, reducing the time and cost to achieve pregnancy.
- Use Case: Opportunistic Screening
Description: The AI checks every pixel of a scan. It might find incidental issues (like early signs of osteoporosis or a lung nodule) on a scan ordered for a completely different reason.