Indica Labs

The cleared product is a platform, not an algorithm. HALO AP Dx is an enterprise digital pathology workspace whose FDA clearance covers primary diagnosis, meaning viewing and diagnosing from digital slides, and that clearance rests on the platform plus scanner pairing rather than on AI performance.

AI is a genuine and central part of the offering, including hosted third party algorithms and the company's own image analysis, but a laboratory could deploy the cleared platform for digital primary diagnosis and use no AI at all. Graded C on the same principle applied to ModMed and Elation Health: an established platform that adds AI is not the same as a product that is AI.

The platform's cleared function is to let a pathologist make the primary diagnosis on a digital slide, so the human is the diagnostic actor by design. Algorithms accessed through the associated store are assistive and, as the company notes, carry their own separate regulatory classifications, which is an important disclosure: platform clearance does not extend to the algorithms running on it. A buyer must evaluate the platform and each algorithm independently.

The regulatory scope is stated with real precision, including clearance numbers tied to named scanner hardware, and the company is consistently explicit about which product is cleared and which is research use only. What is not published is model level detail on its own AI, training data, or validation, because the disclosure emphasis sits on platform clearance and integration rather than algorithm performance. Transparent about scope, quiet about models.

One named regulation does substantive work here and one structural gap is specific to what this company is. The company states built in compliance with the electronic records regulation governing record integrity and audit trails, alongside health privacy rules and, for the research product, European data protection.

That first one matters more than a general security claim for a diagnostic system of record, because it implies audit trail and record integrity controls: a laboratory can establish who accessed and altered what, which is the control that makes a later dispute answerable and which most vendors in this lane do not claim at all. What is absent is the part that follows from the platform's role.

This is an image management platform that hosts algorithms, including third party ones, so it sits in other vendors' supply chains as well as having its own, and nothing published describes monitoring of hosted algorithms, what data a hosted algorithm receives, whether a hosted vendor may retain anything, or what governance applies to admitting one.

This index recorded the same residual on another platform carrying a catalogue of third party applications, and it is the question a marketplace operator uniquely owns. Ask what governs a hosted algorithm's access to images, whether hosted vendors may retain or train on them, and for a sub processor list covering the catalogue.

The 510(k) clearances for primary diagnosis carry validation behind them, and the platform reports deployment at hospitals, contract research organizations, clinical laboratories, and government and academic institutions. But no peer reviewed outcome or concordance literature specific to the platform was located in this review, and the company's public evidence is framed around regulatory milestones and feature releases rather than clinical study results. Cleared and widely deployed, lightly published.

Better documented than most pathology vendors here, though framed as regulatory compliance rather than AI safety specifically. The company states built in compliance with FDA 21 CFR Part 11, which governs electronic records and audit trails, alongside HIPAA and, for the research product, GDPR. Part 11 compliance is meaningful for a diagnostic system of record because it implies audit trail and record integrity controls. What is absent is AI specific safety documentation such as failure mode disclosure or monitoring of hosted algorithms.

HIPAA compliance is stated explicitly and repeatedly as a built in property of both HALO AP Dx and HALO AP, which is more than most vendors in this category publish. A specific BAA commitment or trust documentation was not located, so the posture is asserted compliance rather than a published contractual position, but it is stated plainly by a US headquartered vendor selling into US clinical laboratories.

ISO 13485:2016 certification is stated, which covers medical device quality management rather than information security, and 21 CFR Part 11 compliance addresses electronic record controls. Neither substitutes for SOC 2 or ISO 27001, and no such attestation or trust center was located. More documented than an unrated vendor, short of a security certification.

Cleared for primary diagnosis, which is the highest bar in digital pathology, and cleared twice: K232833 with the Hamamatsu NanoZoomer S360MD in 2024, then K252762 in December 2025 adding the Leica Biosystems Aperio GT 450 DX scanner. CE-IVDR marked in Europe, the UK, and Switzerland.

The critical nuance, which the company states plainly and repeatedly, is that clearance is scanner specific: HALO AP Dx is cleared only with named scanners, and the separate HALO AP product is research use only in the US and explicitly not cleared for diagnostic use. Credit for regulatory achievement and for unusually disciplined disclosure of exactly what is and is not cleared.

The first vendor in this lane to give a concrete answer to the question a platform business raises, and it answers it with product rather than policy. Where a company lets customers build their own models, the governing question is who establishes that a customer built model actually works.

Indica Labs ships validation workflows inside HALO AI as a feature: users train and optimise classifiers, then run them against a validation set, with the result presented as regions of interest marked for true positives, false positives and false negatives, and the whole process centralised rather than assembled by hand.

That is the means of accountability placed in the hands of the party who built the model, which for a platform vendor is the right location for it, and no other platform in this lane offers an equivalent. It is credited as a genuine governance property rather than a convenience feature.

Three things hold it at B. The validation tooling establishes whether a customer's classifier performs on that customer's validation set; it says nothing about performance across patient groups, and no subgroup or bias disclosure was located for Indica's own models or for the platform. There is no responsible AI statement and no AI management system certification.

And the second half of the platform question is untouched: HALO AP hosts third party clinical algorithms, including Ibex Galen, and nothing published addresses who validates a hosted algorithm in a given laboratory, who monitors it after go live, or where accountability sits when a hosted model is wrong. Publishing that division of responsibility, and any subgroup performance, would move this materially.

Scope discipline carries this grade and it is worth crediting because it protects the buyer from an inference they would otherwise make. Regulatory status is stated with real precision, including clearance identifiers tied to named scanner hardware, and the company is consistently explicit about which product is cleared and which is offered for research use only.

Tying a clearance to specific scanners is a form of limitation disclosure in the same family this index credited for enumerated camera compatibility elsewhere: it tells a laboratory that performance is understood to depend on the acquisition device and that the vendor will not claim beyond what was tested. Distinguishing cleared from research use products in your own material also prevents the single clearance from colouring an entire catalogue, which several vendors here allow to happen.

Held at C because the disclosure emphasis sits on platform clearance and integration rather than algorithm performance. No model level detail, training data description or validation for the company's own artificial intelligence was located, so a buyer knows precisely what is authorised and nothing about how well it works, and no warranty, indemnity or remediation commitment attaches. Transparent about scope, quiet about models. Ask for performance on the cleared configurations, what a laboratory must validate itself for the research use products, and what changes when a different scanner is used.

Interoperability is the core of the value proposition rather than a feature. The platform is described as an integration engine bringing together diverse software and hardware, it is cleared across scanners from two different manufacturers, and it hosts third party algorithms through an associated store, making it a hub other vendors plug into.

The Leica clearance specifically covers a DICOM compatible scanner, which matters because DICOM is the standard that makes pathology images portable across systems. Multi vendor hardware clearance plus third party algorithm hosting is the strongest interoperability position available in this category.

The most complete answer to this axis in the lane, because it addresses both halves rather than one. Deployment is presented as a genuine choice between on premise and cloud, and the company treats that choice as consequential rather than incidental.

For cloud, Amazon Web Services is named outright, the deployments are stated to be validated by AWS, and the whole HALO suite is covered including the image analysis platform, the HALO AI classifier add on, the HALO Link image management system and the clinical HALO AP workflow platform. The distinctive property, and the reason this sits at the top of the axis, is that the customer retains full ownership of the AWS account while Indica Labs manages the deployment within it.

That is a materially different arrangement from a vendor hosted service: the infrastructure and therefore the data remain in the customer's own tenancy, which puts region selection and residency under the customer's control rather than requiring a vendor commitment. It is the middle path between an on premise vendor and a software as a service one, and nothing else in this lane offers it.

Implementation is described concretely: a professional services team of certified cloud architects, deployment stated as achievable within days with minimal input from the customer's own IT staff, scaled deployments and sandboxes so capacity matches need, no upfront hardware cost, and on demand start, stop and connect through a web based instance manager. HALO AP integrates into an existing laboratory or hospital information system or runs standalone as a case and image management system.

Named deployments include large reference laboratories, a national cancer institute and a European university pathology institute. What is not published is a residency statement in the conventional sense, naming supported regions or committing to them, which matters for customers who would prefer the vendor to carry that obligation rather than assume it themselves.

No pricing is published for the platform or for algorithms accessed through the store. A buyer should note that total cost includes cleared scanner hardware, the platform, and separately licensed algorithms, and that the scanner specific nature of the clearance means hardware choice and software cost are coupled decisions rather than independent ones.

Broad by construction. Clearance covers primary diagnosis of surgical pathology slides from formalin fixed paraffin embedded tissue generally, not a specific tumor type, so scope spans anatomic pathology rather than one indication. The company also serves life science discovery, translational research, clinical trials, and companion diagnostic development alongside clinical workflows. That span across research and clinical settings, plus tumor board and second opinion workflows, is wider than any single algorithm vendor can offer.