Linvest21

The removal test leaves nothing, since the product is a set of agents rather than software they assist. Specialised agents are named by the role they perform, covering equity analysts, portfolio managers and virtual chief investment officers, and the company describes the platform running end to end from idea generation and research through portfolio management, risk management and performance analysis to customised client communications. A macroeconomic product generates structured institutional analysis in real time. The technology is described as deep domain artificial intelligence rather than general purpose.

One formulation is among the cleanest in this index and is stated as architecture rather than caveat: a human in the loop design in which the models recommend and people decide. That is the correct construction for investment decisions carrying fiduciary weight.

It sits unresolved against the company's own repeated description of the platform as autonomous and end to end, including a virtual chief investment officer agent, and nothing published explains which functions run unsupervised and which require sign off. A vendor cannot be both autonomous and human decided without saying where the line falls, and that is why this is held at B rather than higher.

No accuracy, backtest, validation result or performance attribution is published for agents generating investment recommendations, which is the measurement that matters most here. Structural signals exist in place of it: the founder holds six patents in financial technology and applied artificial intelligence, output is delivered in a structured fifteen section format that is at least inspectable, and the human in the loop design places a professional between the model and the decision. None of that is validation of the models themselves.

The company reports hundreds of billions of dollars in assets under its platform's supervision at Series A stage, which if accurate is substantial adoption, and the round is described as oversubscribed. It was selected as one of twelve emerging companies to present at a major asset management conference in 2025, competing for the show prize. Backing comes from a Silicon Valley venture firm with prior unicorn seed investments. Against that, no institution is named anywhere, the assets figure is self reported without corroboration, and the team stands at around nineteen people.

No boundary statement was located. Agents operating across research, portfolio construction and performance for multiple asset managers observe positioning and decisions that are commercially sensitive between competing firms, and nothing states whether models learn across the client base, whether portfolio data is isolated per tenant, or what a manager contributes by adopting the platform.

No data protection agreement, retention schedule, subprocessor list or deletion commitment was located. The platform generates customised client communications, which means it holds or accesses individual client circumstances alongside portfolio positions, and none of the handling terms are published.

No attestation, certification, trust centre or enumerated framework was located. Institutional asset managers conduct operational due diligence before granting a platform visibility of positions and client records, so review has occurred privately if the reported adoption is accurate, and nothing is published for a prospective firm to begin from.

No regulator, statute, licence or supervisory framework is named. For a platform whose agents generate portfolio recommendations and client communications for regulated asset and wealth managers, that omission is material, since investment advice, suitability and marketing communications are each governed and the buyer's own compliance function would need to know what the platform assumes about them.

No individual is assessed and the adapted exposure runs to end investors whose portfolios are shaped by agent output. Models trained on historical market behaviour and prior allocation decisions reproduce the assumptions embedded in them, and a virtual chief investment officer generating house views at scale concentrates that effect rather than diversifying it, since many firms running the same platform may converge on similar positioning. Nothing addresses model behaviour, correlation risk across clients, or how house views are challenged.

No guarantee, indemnity or correction process was located. The buying firm retains fiduciary responsibility for decisions regardless of what generated the recommendation, which makes an explicit statement of what the vendor stands behind more valuable rather than less, and none exists. The end investor, whose portfolio is constructed and whose communications are drafted by agents, is not addressed at any point.

The technology is characterised as deep domain artificial intelligence, which distinguishes it from general purpose models without identifying what it is. No base model, provider, hosting arrangement or subprocessor appears, and no market data, research or economic data source is named, which matters for a macroeconomic analysis product whose quality depends entirely on the inputs behind it.

No named integration, interface documentation or connected system was located. For a platform claiming end to end coverage from research through portfolio management to client communications, connection into order management, portfolio accounting, custody and client reporting systems is unavoidable in practice, and none of it is described.

No hosting provider, region selection, residency commitment or private deployment option was located. Asset managers handling portfolio positions and client data typically require clarity on where analysis runs, particularly where a platform is described as supervising hundreds of billions in assets.

No pricing, packaging or basis of charge was located. The stated mission of making institutional grade technology affordable for medium and smaller firms is a pricing claim in substance, since affordability is the whole proposition, and no figure, tier or unit accompanies it.

The buyer is identified by role rather than by institution type, covering chief investment officers, portfolio managers, strategists and research leads, which is the right specificity for a platform sold into investment teams, and the stated target is medium and small asset managers rather than the largest firms. Functional coverage spans research, portfolio construction, risk, performance and client reporting across both asset and wealth management. Geographic reach is not described beyond a United States base, and no market or asset class breadth is evidenced.