InvestCloud

A platform trading since 2010 across front, middle and back office, with its model based products described by the company itself as the first of a new generation when they shipped in August 2025. Strip the models and the entire business remains: client portals, adviser workspaces, portfolio reporting, onboarding journeys, managed account infrastructure and the data platform underneath.

Built on the Clearwater and MyComplianceOffice precedent because the models sit inside the regulated operation rather than around it, since Intelligent Screening performs onboarding due diligence and ongoing risk monitoring, which is a regulated obligation rather than an efficiency feature.

Sequencing is stated, if briefly: the models perform research and preparation so that advisers, relationship managers and compliance officers focus on the decision, which places the model upstream of the human judgement rather than in place of it.

That is the same disclosure shape that earned Red Oak a B. Off an A by some distance because no gate, threshold, confidence measure or sampling audit is described for either product, and screening in particular operates a match threshold whether or not the vendor names one.

No accuracy, precision or recall figure, benchmark or validation method was located for either model based product. The gap is most material on screening, where the failure that matters is the hit the system does not raise, a sanctioned party or adverse media match that never reaches a compliance officer and therefore produces no artefact to review.

Published claims are speed and cost framed, faster and more cost effective due diligence, which is the volume and speed for accuracy substitution recorded across this index rather than a measure of correctness.

Named institutions at scale, reported independently rather than only by the vendor, which clears the bar. Off an A because no quantified outcome attaches to any of them: no adoption figure, error reduction, time saving or revenue effect is published for either model based product. One further piece of evidence belongs in a buyer's file and is recorded neutrally.

Independent trade coverage in 2023 reported that the founding executive team departed and that revenue had lagged the valuation set in the 2021 recapitalisation, which is the kind of delivery record a buyer evaluating a multi year platform commitment should weigh alongside the scale figures.

Nothing published describes whether models learn from client content or whether learning is isolated per institution. There is a second and less usual exposure here that follows directly from how the products were built. Both model based products run on technology licensed from third parties, so client onboarding material, screening results and recorded client meeting content pass through supplier systems as well as the vendor's own, and no boundary statement, retention position or supplier data commitment was located for either path. Buying capability rather than building it multiplies the number of parties holding the data and the disclosure has not kept pace.

Encryption of data in transit and at rest is stated for the data platform and nothing further was located: no privacy programme, retention position, data processing terms or subject rights framework. The scope is wide, since the platform holds portfolio positions, transaction history, client interactions and, through the meeting product, the content of recorded conversations between advisers and their clients. Those clients are identified individuals with no relationship to the vendor.

End to end encryption in transit and at rest is named, alongside a general claim of enterprise grade security and governance built into the data platform. No certification, attestation, report type, audit scope, penetration testing summary or trust centre was located. A vendor holding platform records for 27.4 million accounts and named global banks almost certainly holds formal attestations, and none of them are published where a buyer can find them.

A software vendor holding no financial licence of its own, with no supervisory programme, sandbox admission or regulator run assessment of the model layer located. Its clients carry the authorisations and the regulatory obligations the products serve, including onboarding due diligence and ongoing monitoring duties.

No framework, bias testing, fairness evaluation, model documentation or independent assessment was located. The claim made instead is that governed and validated data underpins the models so that they act on reliable inputs rather than ad hoc extracts, which is a data quality argument presented where a governance answer belongs.

Clean inputs do not establish that a screening model treats names, nationalities or entity types even handedly, and adverse media screening is one of the places in this index where that question has the most direct consequence for an individual.

Nothing published describes liability, indemnity or recourse when a model based screening result is wrong in either direction. The recourse question has a specific victim here rather than an abstract one: a screening system that produces a false match attaches an adverse finding to an identified person during onboarding, and nothing describes how that person learns of it, contests it, or has it corrected. The licensed component structure adds a second unanswered question about which party stands behind the determination.

An unusual result on this axis and the reason is structural. The vendor names the third party technology behind both of its model based products, identifying its screening supplier and its adviser assistant supplier by name, which tells a buyer whose systems touch their data. Almost no vendor in this index discloses that much.

Off an A because no base model, provider, version or hosting arrangement is named for any component, and the agentic capabilities announced in the 2026 core platform partnership carry no stated provenance at all. Worth recording as a pattern: the vendors that buy their models disclose a supply chain, and the vendors that build proprietary ones disclose nothing.

Integration depth is the product's central claim and it is evidenced. A unified data model brings client data, interactions and portfolio activity together across front, middle and back office, aggregating externally held accounts alongside trust and advisory positions, and the platform carries 27.4 million accounts.

The May 2026 partnership deploys the front office directly onto a major core processing provider's wealth platforms through that provider's own integration framework, explicitly so institutions can modernise without replacing the systems underneath, which is the same architectural position additiv and Backbase occupy.

Cloud delivered software as a service, with no region choice, hosting location, tenancy model, residency commitment or subprocessor list located. The subprocessor gap is more pointed here than for most vendors in this category, because two model based products are known to run on third party technology and those suppliers are therefore processing client data without any published statement of where.

No pricing, band, module rate or minimum is published. The platform is modular and sold by configuration, which makes the absence more consequential than a missing list price, because the units of purchase and how they combine into a deployment cost are not derivable from anything public. Pricing is by quote following a sales conversation.

Breadth is counted and anchored: more than 150 asset managers, 400 wealth managers and 140,000 individual advisers, supporting 6.5 trillion dollars in assets across 27.4 million accounts, with Wells Fargo, BNP Paribas and Raymond James named in independent trade reporting.

Segments span private banks, wealth managers, registered investment advisers, asset managers, family offices, asset allocators including pension funds and endowments, and asset servicing providers, across North America and Europe. The May 2026 partnership extends reach into an institutional base of more than 600 financial institutions.