Vibrant Practice

The grade describes the mechanism, not the quality. The company is genuinely AI native rather than an established platform with models added, and AI drives the core clinical modules: documentation, laboratory synthesis and treatment plan generation. But what a clinician purchases is a practice operating system, and scheduling, billing, intake and the patient application would remain a coherent product for a functional medicine practice without any model layer at all.

This index applies the same reasoning to every platform whose value stands substantially without the AI, and a C or B here is a factual statement about what is being bought rather than a criticism of the product.

A stated stance without a described mechanism. The founder has publicly argued that human in the loop is the correct approach for both coding and anything relational or patient facing, which is a considered position and better than silence, but this index does not grade a philosophy as an oversight mechanism. No confidence signal, threshold, abstention behaviour or required review step was located.

The autonomy surface also deserves naming. The platform generates personalised treatment plans from patient data, lifestyle factors and history, and pushes personalised recommendations to patients through its own application. Recommendations reaching a patient directly, rather than a clinician who filters them, is a materially different risk surface from a note a clinician reads, and nothing published describes what a clinician approves before a patient sees it. Ask exactly that.

Nothing was located. No model or model family named, no accuracy figure for laboratory interpretation, charting or treatment plan generation, no evaluation methodology and no model card. The company is at pre seed stage and founded recently, so the absence is unremarkable, and it leaves a buyer with nothing to assess.

One feature of the specialty makes disclosure worth more here than the general case, and it is the reason to ask rather than wait. Functional, integrative and longevity practice operates with reference ranges, panels and interpretive frameworks that differ from conventional laboratory reporting, and with a literature base that is uneven in quality. A model generating interpretations and treatment plans in that setting has been shaped by whatever material it learned from, and the range of that material is wider and less settled than in conventional medicine.

So the useful questions are about grounding rather than benchmarks. What does the interpretation layer draw on, is it the ordering laboratory's own reference ranges or something the vendor supplies, are treatment plan suggestions traceable to a source a clinician can inspect, and what happens when sources disagree. A clinician in this field will have their own considered positions, and a system that quietly imports different ones into their charting is changing their practice rather than documenting it.

That is answerable without a model card and without publishing accuracy figures a young company will not have.

Ask which model or models are used, whether an external provider processes clinical content, what the interpretation layer is grounded in, and whether outputs cite their basis.

Nothing identifies any party in the chain: no model or model family, no hosting arrangement and no sub processor list was located, and no retention period, training use statement or de identification posture was found. One published declaration deserves examining rather than passing over.

The patient application's store listing declares that the developer does not collect any data from it, and that is difficult to reconcile with what the application does, since it displays laboratory results, carries secure messaging with a care team, delivers personalised care plans and treatment recommendations, and issues medication and supplement reminders. An application performing those functions is necessarily transmitting and receiving identifiable health information.

There may be a technically defensible reading behind the declaration: the developer may consider that it processes rather than collects, acting on behalf of the practice that holds the record, or that the application is a view onto the practice's data rather than a collection of its own. Both are arguable.

Neither helps the patient reading the listing, who is deciding whether to install something that will hold their laboratory results and is being told nothing is collected, and who will be misled regardless of how the developer reasons about it. This index treats store declarations as vendor authored summaries rather than evidence, and an affirmative statement of none is a stronger claim than an incomplete list. Ask what the application transmits, stores on device and retains, and why the declaration reads as it does.

Commercial validation only. A 1.7 million dollar pre seed round in March 2025 with named institutional investors including Anthemis Group and Emerson Collective is real external diligence, and the founder's product background at Meta is verifiable. Nothing beyond that: no customer is named, no practice count or deployment figure is published, no case study exists and no clinical or operational outcome has been measured. Graded C rather than Not Rated because commercial traction is evidenced; clinical benefit is not.

No retention period, training use statement or de identification posture was located. The earlier assessment asked what the patient application collects, and the second pass produces an answer that raises more questions than it settles.

The application's store listing declares that the developer does not collect any data from it. That is difficult to reconcile with what the application does. It displays laboratory results, carries secure messaging with a care team, delivers personalised care plans and treatment recommendations, and issues medication and supplement reminders. An application performing those functions is necessarily transmitting and receiving identifiable health information.

There may be a technically defensible reading behind the declaration. The developer may consider that it processes rather than collects, acting on behalf of the practice that holds the record, or that the application is a view onto the practice's data rather than a collection of its own. Both are arguable positions.

Neither helps the patient reading the listing. A person deciding whether to install an application that will hold their laboratory results is being told nothing is collected, and that will mislead them regardless of how the developer reasons about it. This index treats store declarations as vendor authored summaries rather than evidence, and an affirmative statement of none is a stronger claim than an incomplete list.

So the questions are: what the application transmits and stores, what is retained on the device, what the platform retains, and whether patient generated content is treated differently from clinician entered content.

Ask for the retention schedule and the training position in contract language, and ask why the declaration reads as it does.

No compliance statement or business associate agreement terms were located, and the point the earlier assessment made is the right frame: as the system of record rather than a tool beside one, this vendor's own posture is the practice's posture. That makes it the first item to establish rather than a later diligence step.

The second pass makes the stakes concrete. At least one established multi site practice has migrated its entire record onto this platform, telling its patients so directly. A migration of that kind moves a practice's complete clinical history, its messaging, its prescribing and its billing into a company founded in 2024 and funded at pre seed scale. That is not an argument against the vendor, and it is the reason the contractual position matters more here than for a point solution: when a system of record vendor fails, is acquired, or changes terms, a practice cannot simply stop using it.

So the questions extend past the agreement itself. Establish what happens to the record on termination, in what format it can be extracted, whether the practice can obtain a complete export including messages and attachments, and what the vendor's obligations are if it ceases operating.

The European dimension the earlier assessment raised stands. Serving practices in both regions means two regimes with different requirements, and the general data protection regulation is not satisfied by measures designed for the United States rule. Establish which entity contracts with European practices and on what basis.

Ask for the agreement, the data portability and termination terms, and the European contracting position.

No attestation and no trust centre were located. The company was founded in 2024 and has raised at pre seed scale, so the absence is unremarkable for its stage and this grade records what a counterparty can verify rather than a judgement that controls are absent.

The earlier assessment's point is why the usual proportionality argument does not fully apply, and the second pass strengthens it. This product is the system of record. It holds the entire clinical history for a practice, not a copy of part of it, alongside prescribing, messaging, scheduling and billing, with a patient facing application on top. A control failure in a point solution exposes what that tool touched. A control failure here exposes everything a practice has.

That asymmetry means the scope of what an examination would cover is larger for this company at its size than for many better funded vendors holding narrower estates. The argument is not that a pre seed company should already hold a report; it is that the substitutes matter more. A clear security page describing hosting, encryption, access control, backup and recovery, subprocessors and the breach process would do most of the practical work for the buyers this product serves, at a fraction of the cost.

Backup and recovery deserve their own mention, since they rarely feature in this index and are central for a system of record. A practice needs to know what happens if data is lost, not only if it is stolen.

The named enterprise style customer migration raises the expectation that some assessment already exists.

Ask what external testing has been performed, and for the recovery and export arrangements.

No clearance, device authorisation or exemption analysis was located. The two capabilities the earlier assessment flagged are confirmed, and the second pass establishes that the patient facing route is the sharper of them.

The application delivers personalised insights and personalised care plans with treatment recommendations directly to patients, alongside medication and supplement reminders. That is the configuration where the reasoning behind the conditional exemption for decision support does not hold. The exemption depends on a professional being able to review the basis of a recommendation and reach an independent judgement. A patient cannot do that. Where a recommendation reaches the patient, whether it was reviewed by their clinician first is the whole question, and nothing published establishes that a clinician approves each item before it appears.

Laboratory interpretation is the second, and in this specialty it carries its own weight. Functional and integrative practice makes extensive use of panels and reference ranges that differ from conventional laboratory reporting, so an interpretation layer is doing more than restating a flagged result.

One further question follows from the setting rather than the software, and it should be asked. Practices in this field frequently dispense supplements, and the application issues supplement reminders. Establish whether the platform recommends specific products, whether any commercial relationship attaches to those recommendations, and whether a clinician selects them. A recommendation that is also a sale is a different object from a clinical instruction.

Ask which framework the company believes governs the treatment planning and the patient facing recommendations, and to see the analysis written down.

The grade describes disclosure. The founder's stated human in the loop position is a genuine credit and unusual to find articulated publicly at this stage. Against that, no fairness, subgroup or demographic performance disclosure of any kind was located, and no governance or responsible AI documentation exists.

One question matters more in this market than in most and is asked here neutrally rather than as a judgement on the field: this index asks every recommendation generating product what its output is GROUNDED IN, and credits products anchored to named published guidance a clinician can trace and dispute. In areas covered by well established society guidelines that question has a clear answer.

In functional, integrative and longevity medicine the guideline base is less standardised and practice patterns vary more widely between clinicians, which makes the grounding question more important rather than less. Nothing published describes what the treatment plan generation is grounded in, and that is the single disclosure that would most improve this record.

Two passes located no model named, no accuracy figure for laboratory interpretation, charting or treatment plan generation, no evaluation methodology, no model card and no warranty, indemnity or remediation commitment. The company is at pre seed stage and recently founded, so the absence is unremarkable in itself, and it leaves a buyer with nothing to assess.

One feature of the specialty makes disclosure worth more here than in the general case, and it is the reason to ask now rather than wait for maturity. Functional, integrative and longevity practice operates with reference ranges, panels and interpretive frameworks that differ from conventional laboratory reporting, and with a literature base that is uneven in quality.

A model generating interpretations and treatment plans in that setting has been shaped by whatever material it learned from, and the range of that material is wider and less settled than in conventional medicine, so two systems could produce confidently different interpretations of the same panel and both would look authoritative. The useful questions are therefore about grounding rather than benchmarks, and they are answerable without a model card.

What does the interpretation layer draw on, is it the ordering laboratory's own reference ranges or something the vendor supplies, are treatment plan suggestions traceable to a source a clinician can inspect, and what happens when sources disagree. A system that quietly imports different positions into a clinician's charting is changing their practice rather than documenting it.

This product IS the electronic health record rather than a tool integrating with one, so the axis inverts: the question is not what it connects to but what can get out of it. Nothing was located on interoperability standards, laboratory interface support, referral exchange or data portability, which matters disproportionately here because functional and longevity practice depends on large volumes of external laboratory data flowing in, and because a clinician replacing their system of record needs to know what happens to the record if they leave. Graded C because the all in one architecture is clearly established and the interoperability picture is entirely absent. Ask specifically about laboratory interfaces and about export on termination.

No hosting model, cloud provider, region or residency commitment was located, and no customer hosted option is offered or implied.

The residency question the earlier assessment raised stands and has a specific shape here. The company sells into North America and Europe, and a single hosted estate serving both raises transfer considerations a United States only vendor does not face. Under the European regime, personal health data is a special category with its own conditions, and transfers outside the region require a lawful mechanism that must be documented rather than assumed. A practice in Europe cannot rely on its vendor's general assurances; it needs to know where the data sits and what mechanism covers any movement.

So establish whether European practices are served from a European region or from the same infrastructure as North American ones, and if the latter, which transfer mechanism applies.

The delivery surface is broader than a web application. There are patient applications on both mobile platforms, including builds for desktop and headset environments, which means clinical content reaches consumer devices outside any practice's control. What is cached on those devices, and whether it persists after logout or uninstall, is part of this axis rather than separate from it.

And because the product is the system of record, the residency answer governs everything the practice holds rather than one workflow. That raises the stakes on getting it in the contract rather than accepting current configuration.

Ask for the hosting region per jurisdiction, the transfer mechanism for European customers, the subprocessor list, and the on device caching behaviour.

No price, tier or figure was located.

The pricing mechanism is disclosed, a subscription model sold to independent clinicians, which is a little more than most vendors in this category offer and is appropriate to a market of small practices that cannot run an enterprise procurement. It is credited but it does not reach a B, because naming a subscription model without any tier or rate tells a buyer almost nothing about the economics, where the vendors graded higher on this axis publish either the tiers themselves or a mechanism that reveals how vendor revenue moves.

Publishing the actual tiers would be straightforward for a product sold to individual clinicians and would place this vendor among the more transparent records in the index rather than among the majority.

Narrow, deliberate and genuinely uncovered elsewhere in this index. The target is independent clinicians practising functional, integrative and longevity medicine, operating in concierge, direct primary care and hybrid business models, across North America and Europe.

That is a coherent and well defined segment with real differences from conventional ambulatory practice: heavier laboratory panels, longer consultations, lifestyle and root cause framing, and a patient population paying directly rather than through insurance. The founder's presence at the Institute for Functional Medicine and A4M indicates the company is engaged with the segment rather than marketing at it. Graded B rather than A because no specialty specific instrument, protocol or panel handling was evidenced, only asserted.