Who it is for

Bank trust departments and independent and state-chartered trust companies. The platform does not replace the trustee of record or the custodian of assets. It is the reasoning layer between them.

Four roles share that layer:

Trust administrators

Where judgment lives. Cited rationales for discretionary work. The officer decides.

Operators

Where arithmetic lives. Fees, principal and income, exceptions — measured against the firm’s definitions.

Compliance teams

Guardrails and audit readiness. Edge cases surface early, with reasoning an examiner can walk.

Investment advisors

Constraints in the same layer as the portfolio — without leaving the workflow.

The work AI should help with

Trust administration is reading what the instrument actually says, applying it to the account in front of you, and leaving a rationale someone else can follow. The work a copilot should take is the slow, repeatable part — not the judgment.

Read

Read governing instruments and trust accounts. Each extracted field carries a citation. Fields it cannot support come back empty, not guessed.

Flag

Flag what needs attention with the reasons cited — exceptions, missing data, a distribution that needs an officer — before they reach a client or examiner.

Prepare

Prepare the file for officer review. The agent produces a proposal. A named person reviews, questions, corrects, and approves.

The controls such institutions need

Fiduciary duties stay with the named officer. The software does not take the appointment, and it does not act on its own. Institutions that answer to examiners need four controls in the architecture — not as a checklist after the fact.

Named human decision

Interpretive and judgment-level actions are routed to a named officer, counsel, or committee. The agent prepares; the human authorizes; the record shows both.

Approval before action

Nothing happens until a person approves. Agents read through the Bridge, propose through the Bridge, and, only after approval, write through the Bridge.

Tamper-evident record

Every decision is written to a tamper-evident record — a hash-chained log an examiner can check later, without relying on anyone's memory. The human decision and the engine's proposal stay distinct.

Overlay, not replacement

The r.Team platform overlays the systems you already run. It does not replace them. Agents run beside the existing system of record until evidence, not hope, authorizes a write.

How this maps to r.Team

A fiduciary copilot on a fiduciary reasoning platform. Three layers. Five services. Keep with the pods. A Trust Ontology so the Reasoning Engine cites instead of guesses.

Three layers

01

Reasoning Engine

Prepares, explains, and proposes beside the team.

02

The Bridge

One doorway to the systems you already run. Read first.

03

The Record

Proves what happened. A fingerprint — never a client file.

Five services

Each one gathers, prepares, watches, or owns the books — and stops for a named person to decide.

Prepare the work and stop

rAgents gather the scattered facts, prepare the file, cite the source, and stop. A named person on your team decides.

rAgent →

A private model for the families you serve

Each Family Trust Account gets its own sealed model, offered under your name. Explains, cites, never advises.

Keep →

A channel the client will answer

Branded, carrier-verified messaging with fiduciary guardrails. Every message is a human decision, written to the permanent record.

Verified Client Messaging →

Watch after a valid login

Watches credentials, privileged roles, service accounts, and vendor sessions. Proposes. Never acts alone.

Inside Guard →

Trust accounting you own

Your data. Your rules. No vendor in between. Maintained by governed rAgents under human approval.

Headless →

Keep with the pods

Family Trust Pod

Keep is offered by the trust company, under the trust company's name, as part of the service a Family Trust Pod provides. One Keep per Family Trust Account.

Trust Administrator Pod

A Trust Administrator Pod attaches alongside Keep for the officer's work; the family never sees the officer's tools, and the officer never sees what the family keeps private.

Trust Ontology

A structured map of the instrument and the securities laws that govern it, so the Reasoning Engine cites instead of guesses. Extends Fiduciary Reasoning — not a service.

Trust Ontology with Securities Law →

Related: What is a fiduciary copilot? · What is fiduciary reasoning? · Platform · Who we serve

Questions teams ask first

Short answers from the published site. Open a question to read it.

Does this replace our trust accounting system?

No. The r.Team platform overlays the systems you already run. It does not replace them. Agents read through the Bridge and write only after a named person approves. Your core system stays the system of record.

Does the software decide?

No. The machine reads. The human rules. Abstracts cite. They do not conclude. The agent prepares; the human authorizes; the record shows both.

What can an examiner check later?

Every decision is written to a tamper-evident record — a hash-chained log of every decision. The human decision and the engine's proposal stay distinct. Change an earlier entry and the break is visible.

Who is it built for?

Bank trust departments and independent and state-chartered trust companies — and the people on the desk: trust administrators, operators, compliance teams, and investment advisors.

What does Keep do for the families we serve?

Keep gives each Family Trust Account its own private model, offered under your name. It explains and cites. It never advises. A Trust Administrator Pod attaches alongside it for the officer's work; the family never sees the officer's tools, and the officer never sees what the family keeps private.

Request a demo

A focused demo of the Reasoning Engine, the Bridge, and the Record — or start a conversation.

A named officer decides, and every decision is recorded.