The governing instrument — the trust agreement that says what the trustee may do — sits in one system. The fee schedule is in another. The custodian, the firm that holds the assets, sent the statement by email. Why a distribution was approved three years ago lives in a retired officer's memory.
An rAgent is a reasoning agent trained by trust professionals and deployed on the r.Team platform. It gathers that information, prepares the work, cites the source, and stops. A named person on your team decides.
Ask a trust operations team where their time goes and the answer is rarely "judgment." It goes to finding things, checking things, and re-doing things — because the facts are scattered.
The same asset into the accounting system, the custodian file, and the tax lot record. A statement address re-typed because mailing instructions changed in one place but not the other. Month-end work that is mostly copying.
"Can this trust make that distribution?" means opening the instrument, finding the article, checking the ascertainable standard — the written limit on what the trustee may distribute, often HEMS (health, education, maintenance, and support) — pulling history, and asking a colleague what was done last time.
The same fee exception is reconciled again. The same clause is re-read when a beneficiary calls. The same break is explained to the auditor because last year's explanation lived in an email nobody kept.
The preparation is done before the officer sits down. The decision is still theirs.
The rAgent is given controlled, read-only access to the instrument, ledgers, statements, and history it needs for the task — and nothing else.
It drafts the setup, the entry, or the rationale in your firm's format, compared against the governing instrument and your own conventions.
Every recommendation points to the clause, the policy, and the record it relied on. An officer, an auditor, or a court can follow it.
Nothing writes back to your systems until that person approves. The proposal and the decision are both kept.
Peered Intelligence — intelligence that works alongside your team at eye level: proposing and explaining, never overriding, with the human decision recorded at every step.
— r.Team product principle
Illustrative workflows — not a catalog, a score, or a performance claim. Which ones are enabled, and at what review tier, is set by your firm.
New holdings, transfers in, and estate assets arriving in a dozen formats.
Custodian and outside-manager statements that must be reconciled before anyone can trust the numbers.
Discretionary requests that depend on the instrument, the standard, and the history.
Billed fees versus the schedule the account is actually on.
The review that takes a week because the inputs are everywhere.
The patterns a busy desk cannot watch for.
Every rAgent runs behind the same four gates. They are not a setting your team has to remember to turn on; they are how the platform is built.
Not every task deserves the same scrutiny. Routine work can be cleared quickly; discretionary work always goes to the officer. The tier is a firm decision, recorded with the workflow.
rAgents never touch your trust accounting, custodian, or document systems directly. They reach them through a single governed doorway — the Bridge — that reads without rewriting and writes only after approval. Every access is logged.
Each proposal and each human decision is written to a per-account, hash-chained record. Change an earlier entry and the break is visible. Years later, "why was this done?" is answered by the record, not by recollection.
Your instruments, accounts, and rationales stay inside your firm's boundary. rAgents are deployed per firm. Nothing about one client's trust informs another firm's work.
r.Team does not ask a trust company to convert anything. The first assignment is to watch — read the instruments, the journals, and the activity logs as your platform actually operates — and report what it finds. Your core system stays the system of record throughout.
The rAgent learns your account codes, your fee conventions, and how your desk really works.
It hands your team a list of recurring exceptions, skipped steps, disagreeing systems, and time sinks — with citations.
Your firm chooses the workflows and the review tiers. Officers start receiving prepared files instead of blank ones.
Hours returned per account, exceptions per thousand transactions, days to close a request — measured against your own books.
A reasoning agent trained by trust professionals and deployed on the r.Team platform. It reads what your team would have had to open, compares it against the governing instrument and your firm's conventions, and hands over a prepared file with every source cited.
No. It never approves a distribution or decides a discretionary act. The officer exercises discretion. The rAgent prepares and stops.
No. r.Team does not ask you to convert anything. rAgents reach your systems through the Bridge. They read without rewriting and write only after a named person approves. Your core system stays the system of record.
No. It does not place a trade, move cash, change a beneficiary, or quietly correct the ledger. It does not send advice to a client or call itself investment advice.
Your instruments, accounts, and rationales stay inside your firm's boundary. rAgents are deployed per firm. Nothing about one client's trust informs another firm's work.
Your firm. Clerical work can be reviewed in a batch; fee adjustments and statement breaks item by item; distributions and anything the instrument leaves to a person, every time. The tier is recorded with the workflow.
Bring a synthetic or redacted account. We will show the prepared file, the citations, and the approval screen — and where the rAgent stops.
Not investment advice. Not a trustee or adviser relationship. rAgent workflows described here are illustrative; availability, review tiers, and scope are set per firm. No customer logos, case studies, pricing, or compliance certifications are implied.
