Tali AI

Ambient generation, dictation and clinical search are all model driven and there is no non AI business underneath. Worth watching rather than downgrading: Tali now describes 14 connected capabilities spanning pre visit preparation, documentation, forms, billing and outbound patient calls, which is platform expansion. If the commercial centre of gravity shifts to workflow breadth, the moat starts to look like the Canadian forms and billing content rather than the model, and this grade should be revisited.

Explicit and well described review gate: the note is generated inside Tali, the clinician reviews and edits it there, and only then chooses to insert it into the patient chart as a deliberate action rather than an automatic write. Separating generation from insertion is a cleaner boundary than products that populate the chart pending review. Held at B because no acceptance rate, edit burden figure or confidence threshold is published.

Technical description is generic, citing natural language processing, medical models and speech recognition without naming any of them, and the performance claims offered are speed rather than accuracy, notably documentation three times faster than typing.

One claim deserves substantiation rather than repetition: Tali states its notes were ranked in the top 1 percent for accuracy and quality by the United States Veterans Affairs, which would be a meaningful independent result but is presented without a date, methodology, comparison set or published source. Ask for the underlying assessment before weighting it.

One published commitment on this axis is precisely worded, and reading it precisely is the whole assessment. The company states it never sells or uses notes or recordings to train third party models. That bounds where content goes externally, which is real and worth crediting, and it leaves entirely open whether customer content trains the company's own models. A buyer skimming the sentence will read a categorical no training promise; the sentence does not say that.

This is the same shape recorded elsewhere in this index where a carefully drawn claim reads broader than it is, and it belongs on this axis because first party training is exactly the question of whether customer content enters a corpus at all. Geography is stated and helps: notes and recordings are both stored and processed within Canada, and access is limited to the clinician and authorised clinic systems, which bounds the chain jurisdictionally. Enumeration is absent.

The technical description is generic, citing language processing, medical models and speech recognition without naming any of them, no foundation model provider or version is identified, no hosting arrangement is published, and no sub processor list was located. Ask whether customer content trains the vendor's own models, get it in the agreement rather than a help page, and ask for the sub processor list.

No peer reviewed publication, controlled evaluation or independent clinical study located. The two external reference points are a procurement vetting rather than clinical evidence, being pre qualification under the Canada Health Infoway AI Scribe Program, and the unsubstantiated Veterans Affairs accuracy ranking noted above. Physician testimonials are named and specific but remain testimonials. Graded C on the standing precedent that vetting and adoption do not substitute for evidence of benefit.

Strong on handling, qualified on training, and the qualification is the reason this is not an A. Data is encrypted in transit and at rest, notes and recordings are both stored and processed within Canada, and access is limited to the clinician and authorised clinic systems. On training, however, the published wording is precise in a way that matters: Tali states it never sells or uses notes or recordings to train third party models.

That commitment leaves first party training unaddressed, which is a materially different promise from the unqualified never used for training statements made by others in this category. Ask directly whether customer content trains Tali's own models, and get the answer in the contract rather than the FAQ.

Multi jurisdiction by design, claiming HIPAA alongside PHIPA and PIPEDA with data residency options, which is broader regulatory coverage than most vendors in this category attempt and reflects a product sold into both Canadian and United States markets. Held at B because business associate agreement terms are not published for inspection before contracting.

The strongest named certification position located in this category. SOC 2 Type II certification is stated consistently across the vendor's own materials and its EMR partners' listings, alongside PHIPA, PIPEDA and HIPAA compliance and configurable data residency. Tali was additionally pre qualified under the Canada Health Infoway AI Scribe Program, which is national body vetting rather than a self declared badge and is the kind of third party assessment almost nothing else here holds. Two things would make it unimpeachable and are worth requesting: the audit date and the scope of the SOC 2 report.

No clearance claimed and none required for ambient documentation. No United States device pathway attaches to a note the clinician reviews and signs, and the vendor states the control point plainly in its own help material: the clinician is responsible for the final patient record and should review and edit before it reaches the chart.

The scope delta here is the product set around the scribe. Alongside dictation and ambient documentation the company ships a medical search function and an assistant that works against the record. Medical search is clinical reference rather than documentation, and it lands in the territory occupied by dedicated clinical reference products rather than by scribes. Establish which modules are in contract at the version being bought.

A less usual question sits in the language support, and it is worth raising because it changes the artefact rather than the classification. The product supports French and Spanish alongside English, and documents Persian as a beta input language that produces English clinical notes. Cross language generation places a translation step inside the documentation pipeline. That is not a device question in the United States, but it does change what the clinician is attesting to when they sign, because the signed record is not in the language the encounter took place in.

Markets are Canada and the United States, the latter served through a separate site path. No United Kingdom or European Union presence was located. Ambient documentation is treated as software as a medical device in those markets, so entry there would require reassessing this position.

A second search confirms the absence and sharpens it rather than repeating it. No fairness statement, no subgroup analysis, no accuracy breakdown by accent, dialect, language or speaker population, and no evaluation methodology were located.

What the second search did find is a claim. The vendor states that the product is designed to recognise a broad variety of accents and that it supports a wide range of voices and accents. That is an assertion about precisely the property this axis asks about, published with no number attached. The shape is familiar across this index: the efficiency gain is quantified, at a stated halving of documentation time and a per encounter time saving, while the robustness claim that would matter most to an underserved patient is asserted and never measured.

The exposure is larger here than for an English only product. The scribe supports French and Spanish, and Persian as a beta input language that produces English clinical notes. Cross language generation inserts a translation step into the record, and no evaluation of it was published. A clinician documenting in a second language has no way to see where the system degrades, and nothing in the workflow surfaces it.

The gap is conspicuous against the rest of this record, because a vendor holding SOC 2 Type II and national programme pre qualification has the compliance maturity to characterise model performance across speaker populations. It is also plainly achievable. Peers in this category have submitted note quality to published evaluation against validated documentation quality instruments and to independent comparative benchmarking, so the methods exist and are already in use in this segment.

The band is reached through the regime rather than through anything the vendor commits to, consistent with how this axis treats the non United States cohort generally. Canadian federal privacy law gives an individual the right to challenge the accuracy of personal information held about them and to have it corrected, and provincial health information legislation carries equivalent correction rights, with regulators behind both.

Processing and storage within Canada keeps the product inside that regime. So the recorded patient has a mechanism, which is the thing this axis exists to test, and it exists whether or not the vendor offers one. On the vendor's own side there is little, and one published claim actively needs substantiating before anyone relies on it. The company states its notes were ranked in the top 1 percent for accuracy and quality by a United States federal health system.

If accurate that would be a rare independent external assessment and the single most valuable evidence on this record, but it is presented with no date, methodology, comparison set or published source, so it cannot be weighed and is not relied on here. The remaining performance claims describe speed rather than correctness. No accuracy figure with a definition, no warranty, no indemnity and no remediation commitment was located. Ask for the underlying assessment behind the ranking claim, and for what the vendor commits to when a note is wrong.

Effectively complete coverage of one national EMR market, which is a different and in some ways harder achievement than partial coverage of several. Integrations span Accuro, TELUS Collaborative Health Record, Med Access, Oscar Pro, PS Suite, Healthquest and Intrahealth Profile, and the integration pattern is launch inside the record with one click rather than a parallel window and paste. Access is also available on desktop, mobile and Chrome for clinicians whose system is not covered. Read the grade as depth within Canada; United States integration depth is asserted rather than evidenced here.

Explicit and unusually complete. Both notes AND recordings are stored and processed within Canada, which is a stronger statement than the note only residency commitments common in this category, and configurable data residency options exist for buyers in other jurisdictions. Stating residency for the audio rather than only the derived text closes the gap most vendors leave open.

A complete published price ladder, which is rare here. A free Personal tier includes 5 ambient scribe sessions and 60 minutes of dictation a month, Premium is CA$45 per month billed annually, Pro is CA$135 per month billed annually, and Enterprise is custom, with a 14 day Pro trial. One structural detail a buyer should notice before anchoring on the mid tier price: ambient scribing, the capability most clinicians are evaluating, is unavailable below Pro. Premium buys unlimited dictation only, so the real entry point for an ambient scribe is CA$135 rather than CA$45.

Deep in Canadian primary care, thin on specialty enumeration. Supports SOAP and DAP note formats, the latter indicating behavioral health workflows, with ICD-10 code suggestions. Its coverage strength is jurisdictional rather than clinical: a fillable forms library auto populating Canadian government, disability and insurance forms, and billing support for Ontario FHO+ time based codes, which is tuned to one provincial primary care funding model. Outbound patient call documentation extends coverage to telephone encounters. No specialty count or language coverage published.