Vise

The removal test leaves nothing to sell. The company's own nonlinear portfolio optimiser is the product, built in house for multi period optimisation that weighs both current and future investment horizons, and it is what produces a personalised portfolio holding individual securities rather than a fund. Everything downstream, rebalancing, tax loss harvesting and trading, is that optimiser running continuously against changing prices and tax positions. The newer intelligence layer adds transformer based reasoning on top. There is no template library or workflow tool underneath that would survive the models being removed.

The position contains a real tension and reading it correctly matters. On advice the company is emphatic, positioning its intelligence layer as assisting rather than replacing advisers, stating that the adviser and client remain in control, and defining itself explicitly against the earlier robo advisory wave that tried to remove the human and largely failed. The intelligence layer suggests changes and drafts the client email; the adviser decides and implements.

On portfolio management the automation is genuine and near total, with earlier company material describing the investment management process as fully automated, and rebalancing, tax loss harvesting and trading all running without a person. That split is coherent rather than contradictory, since the advice relationship stays human while the mechanics do not, but it is the mechanics that move client money, and no threshold, review gate or exception path is described for the automated side.

Two things earn this and neither is a performance record. The architectural choice is stated and reasoned, with the engine described as rooted in advanced statistical models rather than generative artificial intelligence and built around a proprietary nonlinear optimiser, which means the core decision path is deterministic and reproducible in a way a generative system is not, the same argument that supports Recordsure.

And the disclosure discipline around its own numbers is genuinely rigorous, with the value claim marked hypothetical, its tax rate assumptions stated explicitly and the measurement date named, which lets a reader test whether the assumptions hold for them. What is absent is validation: no backtest, no live composite, no error analysis, and no account of how the optimiser behaves in market conditions outside its calibration.

Financial evidence is strong and customer evidence is thin, which is an unusual combination in this index. Revenue reached around 24 million dollars in 2025, roughly 127 to 172 million dollars has been raised across three rounds from 28 investors including several of the best known venture firms, and valuations reported by third parties run between 700 million and one billion dollars.

That the company files public holdings reports confirms it manages real assets rather than only supplying software. Against that, no advisory firm is named as a customer, no assets under management or adviser count is published, and the headline value claim, roughly 1.3 million dollars of additional after tax value on a 4.2 million dollar portfolio over ten years, is explicitly labelled hypothetical and illustrative rather than a client result.

No data boundary statement was located, though the architecture reduces the exposure this axis usually probes. The core engine is a mathematical optimiser rather than a model learning from accumulated client data, so one adviser's portfolios do not train the solver that serves another in the way they would with a machine learning system, and the company is explicit that the intelligence is rooted in statistical models rather than in generative approaches.

The intelligence layer is a different matter, since it is transformer based and ingests client information across an adviser's whole book, and nothing states what it retains, whether that content trains anything, or how it is separated between firms.

No data protection agreement, retention schedule, subprocessor list or deletion commitment was located. The data held is a complete financial profile of each underlying household, since personalising a portfolio requires the client's tax position, realised and unrealised gains, income expectations, concentrated holdings, restrictions and stated values, and the intelligence layer is described as ingesting information on every client in an adviser's book. Registered adviser status brings its own privacy obligations to clients, which is a floor the regulator sets rather than a disclosure the company makes.

No attestation, certification, trust centre or enumerated framework was located. Registered adviser status carries its own supervisory expectations around safeguarding client information and business continuity, which a regulator can examine, and that is a real if indirect floor.

It is not a published assurance artifact, and an advisory firm placing client assets under a third party's discretionary management will be asked by its own compliance function what independent assessment covers the vendor's controls.

The operating entity is a registered investment adviser with the federal securities regulator, stated plainly in its own disclosures, which is a formal admission process passed rather than a compliance claim asserted. Evidenced practice sits alongside the status.

The headline value figure is labelled as hypothetical and illustrative with its assumptions disclosed, naming the realised loss date and the specific long and short term capital gains rates used, and the site carries the required statements that nothing constitutes personalised investment advice, that no offer or solicitation is intended and that past performance does not guarantee future results.

A vendor that publishes the arithmetic behind its own marketing claim is demonstrating the adviser marketing rules working rather than describing them. Seventh A on this axis, on the same basis as QuantumStreet AI.

No consumer credit or identity decision applies, so the axis adapts. The company addresses the attention allocation problem recorded for CogniCor more directly than most, stating that its intelligence layer is designed to ingest information on every client in an adviser's book, from the highest net worth to the lowest, which is an explicit commitment against serving only the largest households. The unaddressed question is whether benefit scales the same way.

Tax loss harvesting and direct indexing generate value in proportion to portfolio size, taxable gains and holding complexity, and the company's own illustration uses a 4.2 million dollar account, so the smallest clients in a book may receive the same attention and materially less advantage. Nothing describes performance or value delivered across account size bands.

No product level guarantee or indemnity was located, and the recourse that exists is stronger than most contractual terms because it is statutory. Operating as a registered investment adviser means the entity carries a fiduciary duty toward the clients whose portfolios it manages, enforceable by a regulator and through the ordinary dispute and arbitration channels, which is a materially harder obligation than the commercial commitments most vendors here decline to make.

Explanation is also built into the product as one of its three stated core functions, so an adviser can account to a client for why a holding changed. What is missing is anything specific to model failure: no correction, notification or remediation process is described for a portfolio outcome traceable to an optimiser error.

The optimiser is stated to be built in house, which shortens the chain at the point that matters most for portfolio construction and is a real disclosure. Everything around it is unnamed. The intelligence layer is described as resting on transformer architecture with no provider identified, no hosting arrangement or subprocessor list appears, and the market and security reference data that any optimiser must consume, including prices, corporate actions, index constituents and tax lot information, comes from suppliers none of which are named.

The platform consolidates rather than integrates, bringing reporting, rebalancing, tax management, trading and model management into one place so an advisory firm can run its portfolio stack from a single system, and the acquisition onboarding proposition depends on that consolidation. Trading and holdings reporting both imply working custodial connectivity, since securities cannot be bought and positions cannot be reported without it. What is not published is the list.

No custodian, portfolio accounting system, financial planning tool or relationship platform is named, and in this market a small number of custodians hold most adviser assets, so a firm cannot confirm its own arrangement is supported without asking.

No hosting provider, region selection, residency commitment or private deployment option was located. Exposure is lower than for a multinational vendor since the business, its clients and its regulator are all in one jurisdiction, so cross border transfer is unlikely to arise, but that is an inference from footprint rather than a commitment. Nothing states where household financial profiles and portfolio records rest.

No pricing is published on the company's own material and the basis of charge is not stated, which for an asset manager would normally be a percentage of assets. One point is worth a buyer's attention: because the entity is a registered investment adviser, its fee schedule, conflicts and business practices are set out in a public regulatory disclosure brochure that anyone can retrieve from the regulator's own database. The information therefore exists in a filed public document even though the vendor does not market it, which is a materially better position than a vendor whose economics are simply unknowable.

The buyer set is independent financial advisers and registered advisory firms, with a specific proposition for firms consolidating acquisitions, where the platform is presented as unifying investment decisions across advisers and onboarding an acquired book in weeks rather than months.

Portfolio coverage is broad within its lane, spanning separately managed accounts, direct indexing that holds index constituents rather than the fund wrapper, personalisation for values and restrictions, concentrated position management, and from late 2025 a customisable long short strategy. The limits are that this is portfolio management rather than the wider advisory business, and no market outside the United States is evidenced.