VieCure

Artificial intelligence is claimed and the engine described is rules. The home page calls Halo Intelligence an artificial intelligence platform and a neural operating system for cancer providers, and the platform page names an AI Inference Engine that works by inference from patient specific data and codified rules, drawing on a knowledge base of more than 5,000 clinical rules covering plans, pathways, protocols, drug and toxicity management, payer rules, and trials and eligibility criteria.

Nothing published describes a model whose parameters are estimated from data. Rules based inference over curated oncology guidance is a defensible design, and what the platform sells without any learned component is substantial: an oncology medical record, a practice operating system, physician and patient applications, and a curated knowledge base that includes NCCN template based treatment plans.

The system proposes and a clinician decides, and the controls around that are asserted rather than described. Halo Intelligence generates, manages and monitors personalized treatment plans, and the company describes blending the provider's judgment and its knowledge base with its inference engine; the oncologist works from the plan rather than the system ordering treatment. Two parts of the oversight are missing.

Nothing says what the system does when a patient falls outside the codified pathways, which is common in oncology, or when licensed sources disagree. And nothing describes what happens after a recommendation is wrong: how overrides are recorded, how rule errors are reported and corrected, or how quickly the library changes when a guideline does.

The inference design is public, and the update discipline is not. The platform page describes the AI Inference Engine as inference from patient specific data and codified rules, over a knowledge base of more than 5,000 rules covering clinical plans, pathways, protocols, drug and toxicity management, payer rules, and trials and eligibility criteria, and part of the content is NCCN template based treatment plans under a 2025 licensing agreement. That explains what produces a recommendation.

The home page's description of a neural operating system does not match that design and is not explained. Nothing says who authors and reviews rules outside the licensed content, how licensed sources are reconciled where they diverge, or how rule changes are versioned and communicated to practices.

The chain behind a recommendation is short, and the rest of the party list is not published. Recommendations come from VieCure's own AI Inference Engine applying its codified rule base to patient data, with part of the content licensed from the National Comprehensive Cancer Network, so no outside model provider sits between a patient record and a recommendation.

Everything else is unnamed: no hosting provider or subprocessor list is published, third party integrations are described by function only (diagnostics, pharmacy, scheduling, billing, radiation therapy, lab), and the website privacy policy refers to trusted partners without naming them. For a medical record that also runs an analytics environment and a patient mobile application, the hosting and analytics parties matter for breach scope.

Named customers at scale, and no published outcomes. The company reports more than 18,000 patients managed daily across more than 100 community cancer center locations, with named users including American Oncology Network, The Oncology Institute, Oncology Care Partners and Summit Cancer Centers, and Commonwealth Cancer Center announced its adoption.

No peer reviewed study, guideline concordance rate before and after adoption, variability reduction figure, or outcome comparison is published. For a product whose central claim is reducing unwarranted variation in cancer treatment, guideline concordance is the obvious measurable endpoint and it is not reported, and scale of use is not evidence of benefit.

The published privacy policy does not reach patient data, and the questions the platform raises are unanswered. The policy (VieCure Inc, effective 4 March 2021) describes the website as an e-commerce site, uses billing and credit card data as its example, and cites SSL encryption and a VeriSign authentication badge; it says nothing about protected health information, patient data, de-identification or research use. The company's site also publishes terms of use and API terms of use.

This is an oncology medical record reported to manage more than 18,000 patients daily, with an analytics environment over its data and a patient mobile application for outcomes reporting and remote monitoring, so retention, secondary use of de-identified data, and who may access curated datasets are the questions a practice needs answered. Safety engineering, meaning how a wrong recommendation is detected and handled, is not published.

No business associate position appears in the published privacy material, and the company's other published terms are where it would sit. The privacy policy (effective 4 March 2021) covers the website as an e-commerce site and does not mention HIPAA or patient data; the site also publishes terms of use and API terms of use.

VieCure sells the oncology medical record itself, so it holds protected health information for every practice that uses it, which makes it a business associate that must sign an agreement with each practice. The buyer here is usually a community oncology practice without a dedicated privacy office, the buyer most helped by finding the terms published.

Federal health IT certification is the outside check published, and no attestation of how the company runs security is. The company publishes ONC certification criteria for its health IT and mandatory certification disclosures referring to its certified capabilities; that certification tests software features, including privacy and security functions, rather than the vendor's operations.

No SOC 2 report, HITRUST or ISO/IEC 27001 certification, trust page, penetration testing statement or vulnerability disclosure policy is published, and the only security language in the privacy policy is SSL encryption and a VeriSign authentication badge on the website. For a cloud medical record holding complete oncology charts across more than 100 locations, provider customers commonly require an operational attestation.

No device claim is made and no regulatory position is stated for software that generates treatment plans. Halo Intelligence produces personalized cancer treatment plans at the point of care, which is the territory the clinical decision support provisions of the 21st Century Cures Act were written for: software stays outside device regulation where a clinician can independently review the basis of a recommendation rather than relying on it.

The rules based design, with NCCN template based plans as part of the content, gives the company that argument, and nothing published makes it or says whether the clinician sees the rule and source behind each recommendation.

The company publishes certification disclosures and nothing that describes how the rule base is governed or tested. The knowledge base of more than 5,000 codified rules includes NCCN template based treatment plans under a 2025 licensing agreement, so part of the content is traceable to a published guideline body; who authors and reviews the other rules, what evidence standard admits one, and how quickly rules change when guidance changes are not described, and no equity or subgroup analysis is published.

Two exposures sit underneath. Guidelines rest on trial populations that underrepresent Black, Hispanic and older patients, and a system that applies them faithfully carries that gap consistently and at scale. And the rule base mixes payer rules with clinical plans and pathways, so a clinician needs to see whether an option is absent for clinical or coverage reasons.

A clinician decides on each plan and can check part of its basis, and nothing stands behind the output. Halo Intelligence generates, manages and monitors personalized treatment plans that the oncologist works from, and the company describes the platform as blending the provider's judgment and its knowledge base with its inference engine, so a plan passes through a clinician before treatment is ordered.

Part of the content comes from NCCN template based treatment plans, a published source a clinician can read and disagree with. No accuracy figure, warranty, indemnity or remedy for a wrong recommendation is published, and nothing says how the platform tells a clinician when licensed sources diverge or when a patient falls outside the codified pathways, which is common in oncology.

The platform is the oncology record itself, with certified capabilities and integrations named by function. Halo Intelligence combines the medical record, treatment planning, imaging, labs, next generation sequencing and molecular pathology in one system, and ingests discrete data from outside sources such as sequencing results. Third party integrations are listed by function: diagnostics ordering and results, pharmacy management, scheduling, billing, radiation therapy and lab management.

The company publishes certification disclosures stating no data portability limits and no contractual limits on its certified capabilities, publishes API terms of use, and lists ONC certification criteria, with deployments at named community oncology networks behind it. No external record system, laboratory or sequencing partner is named, so a practice inside a larger health system cannot tell how Halo coexists with the enterprise record.

Nothing published says where the platform runs or where patient data rests. The company describes Halo Intelligence as cloud based, reached through the web and through physician and patient mobile applications, and its certification disclosures state no workstation, license or bandwidth limits on the certified capabilities.

The Get Halo page describes the product as analyzing patient data and clinical and financial guidelines to deliver personalized treatment options, and offers a custom financial analysis, without any deployment option, hosting provider, region or data location. The terms of use (Viviphi Ltd d/b/a VieCure, effective 27 May 2020) cover the website as an e-commerce site, and the privacy policy covers only the website.

No hosting provider, region, tenancy model, residency commitment or single tenant option is published. For a product acting as the system of record for a practice's oncology population, where patient records and the analytics environment rest matters.

No rate is published, and the fee structure is. The company's certification disclosures describe transaction fees charged as recurring monthly or quarterly charges and implementation fees charged once, with pricing that depends on clinic volumes, practice size, and the scope and complexity of interface requirements, and state that contracts place no limits on using the certified capabilities. That lays out how the product is sold and what drives the cost, but not enough to size it. Nothing defines what counts as a transaction, which decides how cost grows for a network adding practices, or what the implementation fee covers.

Deliberately narrow and named, with deployments behind the core of it. The target is community oncology practices, spanning medical, radiation and surgical oncology from diagnosis through therapy and survivorship, at more than 100 locations across named networks. The patient mobile application extends the platform to outcomes reporting, remote patient and therapy monitoring, toxicity assessments and medication management between visits, and tele-oncology is offered.

Nothing published validates the remote monitoring or toxicity assessment functions, or shows results by discipline. The boundaries are a single disease area, ambulatory rather than inpatient care, and United States practices, and the platform is a record system commitment rather than an add on.