Harrison.ai
Australian AI that reads 124+ chest X-ray findings in under a minute
Overview
Harrison.ai is a Sydney-based healthcare AI company founded in 2018 by brothers Dr. Aengus Tran and Dimitry Tran. The company builds AI-powered diagnostic support tools for radiology and pathology, operating under two product lines: Harrison.ai Radiology (formerly Annalise.ai, a joint venture with I-MED Radiology acquired in full in December 2024) and Franklin.ai (a pathology AI joint venture with global diagnostics leader Sonic Healthcare). Its flagship radiology products can detect up to 124 chest X-ray findings and up to 130 non-contrast head CT findings in under a minute, and are clinically deployed at over 1,000 sites across 40-plus countries — including every public emergency department in Hong Kong and over half of Australia's radiologists.
Beyond its cleared software-as-medical-device products, Harrison.ai has developed Harrison.Rad 1.5, a radiology-specific multimodal foundation model capable of drafting full reports from images, prior studies, and clinical context — the first AI model to pass the UK FRCR 2B Short Case exam, the standard used to certify British radiologists. In November 2025, the company launched its Open Platform, a zero-markup infrastructure layer that lets hospitals integrate once via PACS, RIS, and EHR connectors and then deploy Harrison.ai's native algorithms alongside third-party AI vendors without the 30–60% platform markups typical in the industry. The company has raised over US$240 million in total, including a US$112 million Series C in February 2025.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 11 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Chat access at chat.harrison.ai
- Non-clinical, research-only use
- Pay compute and support at cost only
- No 30-60% platform markup
- Single integration, multiple AI vendors
- Regulatory-cleared products (Annalise, Franklin)
- Priced per deployment — contact sales
Use cases
What you can produce with Harrison.ai
- Automated worklist prioritisation flags for time-critical chest X-ray and head CT findings delivered within one minute of image acquisition
- Detection output covering up to 124 chest X-ray findings and up to 130 non-contrast head CT findings per study
- AI-drafted radiology report for review and sign-off by a radiologist (Harrison.Rad 1.5, research/API access)
- Single PACS/RIS/EHR integration point enabling access to Harrison.ai native algorithms and approved third-party AI vendors via the Open Platform
- Medicare NTAP-eligible AI triage decision for obstructive hydrocephalus in qualifying US inpatient encounters (up to $241.39 reimbursement per case)
- Prostate biopsy digital pathology analysis output to support histopathologist review (Franklin.ai, currently research use only)
ASEAN Perspective
Harrison.ai in Southeast Asia
Harrison.ai has meaningful and verified ASEAN deployments: all public emergency departments in Hong Kong use its radiology AI, Ramsay Sime Darby Health Care in Malaysia deployed Annalise Enterprise CXR across flagship hospitals (Subang Jaya Medical Centre and Bukit Tinggi Medical Centre) using AWS infrastructure, and the company holds regulatory clearance in Singapore, Malaysia, India, and Vietnam. The company is actively hiring a Singapore-based Senior Product Manager for integrations, signalling a deepening regional commercial push. However, data residency practices for ASEAN customers are not publicly disclosed in granular detail — healthcare buyers in Singapore (PDPA), Malaysia (PDPA 2010), and Indonesia (GR 71/2019) will need to verify whether patient imaging data is processed in-country or routed through offshore AWS regions, which remains a critical compliance question for public-sector hospital procurement in these markets.
Harrison.ai is one of the most clinically validated medical imaging AI platforms globally, backed by peer-reviewed publications in The Lancet Digital Health and European Radiology, 12 FDA clearances, three FDA Breakthrough device designations, and Medicare NTAP reimbursement for obstructive hydrocephalus detection. Its Annalise-lineage products cover a uniquely wide range of findings (up to 124 on CXR and 130 on head CT), and the November 2025 Open Platform move — eliminating the 30–60% markup that other platform vendors charge — represents a genuinely disruptive pricing signal for health systems evaluating multi-vendor AI stacks. The Harrison.Rad 1.5 foundation model, which passed the UK FRCR 2B board exam in June 2026, points toward automated report drafting as a credible near-term commercial product.
Caveats are real: the core radiology and pathology products are sold only via enterprise contracts with no public list pricing, making cost assessment opaque for smaller health systems or individual hospitals. Franklin.ai's prostate product was still labelled "Research Use Only" as of mid-2025, limiting its clinical deployment. The Harrison.Rad 1.5 model is not yet cleared for clinical use in any jurisdiction. ASEAN deployment is meaningful but concentrated in private hospital networks (Malaysia's Ramsay Sime Darby) and the Hong Kong public sector, with limited disclosed traction in Indonesia, Thailand, or the Philippines. Data residency arrangements for ASEAN customers are structured via AWS regional infrastructure but are not publicly documented in detail.
What people say
Nine in ten radiologists who trialed Harrison.ai's chest X-ray tool said their attitude toward AI improved after using it, and every participant in that same pilot rated the interface easy to fold into existing reporting workflows — rare numbers for a device category that usually meets resistance. Independent validation backs up the anecdote: a Lancet Digital Health study measured a 45% accuracy lift and 12% faster reporting when Annalise Enterprise CXR augmented radiologist reads, and Harrison.Rad 1.5 became the first AI system to clear the UK's FRCR 2B Short Case exam, the board test used to certify British radiologists, with a median score of 86.5 against a 73.2 pass mark.
Public review volume is thin here — this is enterprise medical software sold through direct contracts, not a G2 or Capterra category — so there's no crowd-sourced star rating to point to. What surfaces instead is deployment and regulatory evidence: over 1,000 sites across 40-plus countries, every public emergency department in Hong Kong, and a Mass General Brigham partnership. A September 2024 privacy inquiry into I-MED's data-sharing arrangement with Annalise.ai was closed by Australia's OAIC without action after regulators found the shared scans sufficiently de-identified.
The catch for buyers: pricing is opaque by design and quoted only through sales. The November 2025 Open Platform move — eliminating the 30-60% integration markup rival platforms charge — is a genuinely disruptive pricing signal, but health systems still need a procurement conversation to find out what they'd actually pay.
Summary of public user & expert reviews, compiled by RECATOOLS.
Notable facts
- Annalise.ai was co-developed with I-MED Radiology and achieved its first regulatory clearance in under 18 months of inception — unusually fast for a medical-device AI.
- Harrison.Rad 1.5 (released June 2026) is the first AI model to pass the UK FRCR 2B Short Case exam, the same standard used to certify practising radiologists in the UK.
- The company was founded by two Vietnamese-Australian brothers; CEO Dr. Aengus Tran was a practising doctor who observed diagnostic capacity constraints firsthand before pivoting to AI.
- Harrison.ai's Open Platform (launched Nov 2025) accepts competing AI algorithms — including direct rivals — as long as they meet standard containerisation specs, an unusually open competitive stance for enterprise medical software.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including Harrison.ai's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Harrison.ai unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Harrison.ai directly →
Spotted something out of date? Suggest an update →
Alternatives to Harrison.ai
More in Healthcare
