Zocks

The models are the product rather than a layer above one. Founded in 2022, the company has never had a pre inference version of itself: the entire proposition is converting spoken client conversation into structured records, and every downstream function depends on that conversion working. Remove the inference and nothing is left, because there is no recording to fall back on and no manually maintained database underneath.

The agentic layer extends the same dependency across an adviser's whole book, answering questions about the client base and proposing actions, and the company describes a multi agent architecture with separate agents assigned to particular workflows. This is the shape this axis is designed to identify.

The oversight design is published and it sits at a sensible point. The agentic layer proposes rather than acts: it surfaces an opportunity across the book, suggests a next action, and the adviser completes that step with a deliberate click, which places a human decision between the model's judgement and anything that reaches a client.

The company states plainly that the agents operate with the user in full control, and enterprise deployments carry fine grained access controls and a named supervision function, which is a specific regulatory concept in this market rather than a general word.

Held below the top grade because the drafting outputs are less clearly gated than the agentic ones: client emails and completed intake forms are generated automatically and nothing published describes a review step before they leave, or whether generated text is marked as such.

The accuracy claim is superlative and unmeasured, which is a combination worth naming rather than passing over. The company states it sets the industry standard with unmatched accuracy and speaker attribution in complex multi participant conversations, and publishes no figure behind it: no word or attribution error rate, no benchmark against any named alternative, no evaluation method, no drift statement.

The claim is also unusually consequential, because attribution decides which participant in a family meeting is recorded as having said a thing, and the note becomes the firm's record of a conversation that was never recorded and therefore cannot be replayed to check.

Several enterprise customers are named openly and they are large, checkable institutions rather than descriptions of a firm type, covering broker dealer networks, adviser aggregators and an insurance group, and one further wealth manager selection was separately announced. A stated base of more than five thousand financial firms gives scale.

The quantified claim, that advisers save more than ten hours a week on administrative work, is repeated consistently across funding announcements and product material and is entirely self reported. The named institutions and the number never join: no case study attaches a measured result to any identified firm, which is what separates this grade from the one above it.

Ownership and control of data are asserted repeatedly and the training question is left open. The company states that customers own and control their data and decide retention, which addresses possession rather than use, and this index treats those as separate questions.

Searched for a statement that client conversation content is not used to train, tune or evaluate models, for any account of whether one firm's material can influence output shown to another, and for terms covering what a model provider retains, and located none of it. A direct denial would be a strong disclosure here and its absence is notable, because a competitor in the same market publishes exactly that sentence.

This is the one axis where the vendor has done something structural rather than said something reassuring. The platform captures and attributes multi participant client conversations without making recordings, which is an architectural decision that removes an entire category of sensitive material from existence rather than protecting it after the fact, and it is stated consistently across product pages and every announcement rather than appearing once in marketing.

Retention is described as the customer's decision, covering how much is kept and for how long, which places control with the regulated firm that carries the obligation. Held below the top grade because the corresponding detail is missing: what structured data is retained once the conversation is gone, a published retention schedule, a deletion commitment and a data processing statement were all searched for and not located.

Recorded as an unverified absence, and the language encountered is the reason it stays here. Security is described through adjectives rather than credentials: enterprise grade, enterprise ready controls, compliance safeguards, and a claim to lead the industry in data security. Not one names a standard, an audit report type or a certificate.

This is the fourth vendor in this sweep session to reach for a grade adjective where an established firm would name a certification, and an adjective earns nothing on this axis. A dedicated security or trust page was not located and read, so the queued check stands and would likely settle it, since a company selling to large broker dealer networks will have been through their vendor reviews.

A technology supplier of this kind holds no licence and requires none, and no penalty applies for that. The downstream question is live and only partly addressed. Client meeting records and the notes derived from them are books and records for a regulated advisory firm, subject to retention obligations and to examination, and the company does name a supervision capability for enterprise firms, which acknowledges the obligation exists.

What is absent is the specification: whether records produced by the platform satisfy retention format requirements, how a firm evidences the completeness of a note produced from a conversation that was never recorded, and what an examiner would be shown.

Two distinct exposures sit here and neither is addressed. The first is transcription and attribution performance across accents, dialects and speech patterns, which is where systems of this kind are known to vary and where a weaker result means a worse record for the client concerned.

The second is the opportunity engine: a model that scans a book of business and surfaces which clients merit a next action is allocating adviser attention, and any systematic tendency in that ranking directs service toward some clients and away from others. Searched for evaluation across speaker populations, or any account of how opportunities are ranked and whether that ranking was examined, and located none.

Searched for a warranty, an accuracy commitment, a service level, a correction obligation or any allocation of responsibility, and located none, and no disclaimer placing it plainly on the advisory firm was located either. The exposure is specific to the design: because no recording exists, a note that misstates what a client said cannot be checked against the source, and the note is the firm's record.

If a dispute later turns on what a client instructed, the firm holds a machine generated account of a conversation with nothing behind it, and the published material says nothing about who carries that.

The stack is described as proprietary and the suppliers behind it are not named. Searched product pages, funding announcements and press material for a provider, a model family, a version or an inference host, and located none. One published detail invites a mistake worth flagging: the protocol server names two widely used general purpose assistants a firm can connect from, and those are the buyer's own tools rather than the vendor's suppliers.

Naming which assistant may reach the data says nothing about whose model listens to a client conversation and writes the note, which for a company whose entire product is that transformation is the disclosure that matters most.

Integration is the mechanism rather than an accessory, because structured conversation data is only useful once it lands in the systems an adviser already runs. Three routes are published: out of the box connectors, a programmatic interface, and an interoperability protocol server that also reaches the general purpose assistants a firm may already use.

The connections are described as two way, and the strongest claim is a write capability, with conversation data automatically populating financial planning, onboarding and proposal systems rather than merely reading from them. Held below the top grade because the systems are named by category and not individually: no client relationship, planning or proposal product is identified, and in this market the specific names are what a buyer is checking for.

Recorded as an unverified absence. The platform is delivered as hosted software and nothing about topology is published. Searched for hosting regions, storage location, a tenancy statement, a private deployment option for enterprise firms and any contractual residency commitment, and located none.

The protocol server raises the sharper version of the question and it is unanswered: when a firm reaches its own client conversation data from a general purpose assistant, that data crosses into an environment the vendor does not control, and nothing states where the resulting processing happens or whose terms then govern it.

A free trial is offered, which lowers the barrier to evaluation without disclosing anything about cost. Searched for a rate, a per seat or per adviser figure, a tier structure or any statement of what moves the price between a solo practitioner and an enterprise deployment, and located none.

The product is explicitly sold across a range from individual advisers to large firms, which is the structure where per seat economics matter most and where a small practice most needs to know before booking a call.

The stated range runs from solo practitioners through mid sized practices to enterprise deployments, with more than five thousand financial firms claimed and named enterprise customers spanning independent broker dealer networks, registered investment adviser aggregators and a mutual insurance and annuity group, which is a genuine spread of institution types within advice.

Enterprise capability is described in terms that match how these firms are actually organised, with organisational hierarchies, fine grained access control and supervision at firm level rather than a single tier product. What holds this below the top grade is geography and adjacency: the footprint appears to be the United States alone, and while investors describe ambitions across wealth, banking and insurance, the published customer base sits almost entirely in advice.