Vocabulary

Composable

A system assembled from independent building blocks rather than bought as one monolith. Blocks can be added, removed, or rearranged as your firm's needs change — without a forced migration or an all-or-nothing vendor bundle.

Block

The smallest independently useful unit in r.team. Every block has a clean interface and works without requiring any other block. You can deploy one, prove it out, and compose the rest on your timeline.

Private inference model block

A frontier-class AI model running privately for one firm. Prompts and data stay inside your boundary — your information never trains someone else's model. Each block exposes a clean interface so it can be swapped or upgraded without touching anything downstream.

Agent pod (idea pod)

A task-focused agent packaged as an independent unit — research, drafting, reconciliation, monitoring. Each pod does one job well and hands work to the next through defined interfaces, not ad hoc scripts.

Stackable

Blocks and pods combine vertically into pipelines. The output of one is a valid input to the next, so you can build workflows piece by piece — the same way you assemble a portfolio from individual holdings.

Event-triggerable

Blocks and pods activate on real-world events — a custodian file arrives, a client record changes, a threshold trips — rather than on schedules or manual runs. They run their job and go quiet. No polling, no batch windows, no babysitting.

Interoperable

Connects to the CRM, custodians, and reporting your firm already runs. r.team fits your existing stack rather than replacing it. Each block talks to the others through clean interfaces, not proprietary lock-in.

Inference

In AI and machine learning, inference is what happens when a trained model runs on new data to produce an output — an answer, a classification, a flag, or a draft. It is not human guesswork, logical deduction in the everyday sense, or regulatory "inference."

When r.team says inference, we mean governed AI producing results from your firm's operational data inside your boundary — the step where a private model block or agent pod turns connected custodian, fee, and account signals into something your team can review and act on.

Ontology

A structured model of the entities in your domain and how they relate — accounts, clients, beneficiaries, fees, holdings, conventions, and the links between them. Your ontology is the map that AI and agents analyze.

r.team connects custodian feeds, CRM records, and trust accounting data into an ontology your operators already trust, so automated analysis stays grounded in fiduciary reality instead of floating free of your systems of record.

Frontier intelligence harnessing

Putting the most capable current AI to work inside your own controls, governance, and data boundary. Frontier-class capability — privately, on your terms, one block at a time.

Monolith

The all-or-nothing platform model r.team is the alternative to: one suite, one migration, one vendor's roadmap. You adopt the whole package — including the parts you do not want — and bend your workflow to fit theirs.