
Intelligence · platform add-on
Every rival says AI. Here is exactly what ours does.
Four specific jobs, each with the inputs it reads, the output it produces, and what it will not do.
What this is
This is an analytics layer, not an adviser. It surfaces and explains. Your judgement decides.
Nothing here places a trade, contacts a client or makes a recommendation in your name. It reads the book you already have and tells you what it notices, with the reasoning attached so you can disagree with it. A tool that cannot show its working is not usable in a regulated advice conversation, and we would rather build the one that can.
The four capabilities
What each one reads, produces, and refuses to do.
Portfolio drift detection
Reads: target allocation, live positions, valuation history. Produces: which portfolios have moved off their target, ranked by how much it matters rather than alphabetically. Will not: decide whether the drift should be corrected.
Next best action for relationship managers
Reads: portfolio state, interaction history, upcoming expiries and failed mandates. Produces: which client to call today and the sentence explaining why. Will not: contact the client, or rank clients by revenue alone.
Rebalancing suggestions
Reads: current allocation, target allocation, holding periods and cost basis. Produces: the trades that close the gap, with the tax consequence attached to each. Will not: place the trades, or ignore the tax cost to make the allocation look tidier.
Explainable recommendations
Reads: everything the other three read. Produces: the reasoning chain behind each suggestion, exportable into the client file. Will not: give you a conclusion you cannot audit or defend to a client.
The category
AI in this market, stated plainly.
| Finvica | Vendor A | Vendor B | Vendor C | |
|---|---|---|---|---|
| What is claimed | Four named jobs, each described | AI-based | AI-powered research | A named AI persona |
| What is explained | Inputs, outputs and limits, per capability | Not stated | Not stated | Positioning, not mechanism |
| Whether the reasoning is auditable | Yes — the chain exports into the client file | Not published | Not published | Not published |
| Whether it advises | No — it surfaces and explains | Not published | Not published | Not published |
Competitors are unnamed here until each row is sourced and dated to the standard 03 §5 requires. Naming a rival and then describing their product from memory is exactly the sloppiness this page exists to contrast with.
Data handling
What the model sees, and what it never keeps.
The analytics layer reads your firm’s own book — positions, valuations, allocations, interaction history. It does not read another firm’s book, and your data is not pooled with anyone else’s.
Client data does not train models. Your book answers questions about your book. It does not become training data, it does not improve a shared system, and it does not leave the boundary your tenant sits in.
Where inference runs, what is retained and for how long are stated in full on the security page alongside the rest of the data-handling position, rather than summarised here where the detail would be lost.
Packaging
How it is offered.
Intelligence is a platform add-on rather than a tier. Scope and commercials are agreed on the demo call, on the same model as everything else: one per-firm subscription, never linked to your AUM, and no share of your revenue.
Questions about the analytics.
Is this required?
No. It is a layer on top of the platform, and the platform is complete without it. If you do not want analytics making suggestions, turn it off and nothing else changes.
Does my client data train your models?
No. Your clients’ data is used to answer questions about your clients’ portfolios and for nothing else. It does not train a shared model, and it is not pooled with any other firm’s book.
Can I audit a recommendation?
Yes, and that is the point of the fourth capability. Every suggestion carries the reasoning that produced it, and that chain exports into the client file — which is what makes it defensible in a review.
Which plans include it?
It is offered as a platform add-on. Scope and commercials are discussed on the demo call, on the same model as everything else — you pay for the platform, never a share of your book.
Can I turn it off?
Yes, per firm and per user. A relationship manager who does not want suggestions does not get them.