When an Advisor Leaves, What Walks Out the Door With Them? 

Ask a firm what it loses when an advisor leaves, and the answer is usually a number: assets, clients, revenue. Ask the person who inherits the book, and the answer is a set of records they cannot yet read. 

Everything is there.  The households, the accounts, the notes going back years, the task somebody opened last spring. What is missing is the thing that made all of it legible, like knowing which of those facts mattered, which had already been raised and set aside, and who in each household decides. 

That knowledge was never anyone’s job to write down. It lived in the head of the person who built the relationship. 

Why a handover document is not the answer 

The obvious response is documentation. Before the departure, the outgoing advisor writes up each relationship and the successor reads it. 

That helps, and it does not solve the problem, for a reason that has nothing to do with effort. A handover document is a snapshot. It is accurate on the day it is written and starts ageing immediately, while the relationship it describes keeps moving. Six months on, a successor can be reading a careful description of a household that has since had a death, a liquidity event, or a change of trustee. 

The context is also not in one place to gather. In Advisor360’s 2026 Connected Wealth Report, a survey of 300 advisors, 74% said their firm’s technology lacks proper integration. When the history of a relationship is spread across a CRM, a planning tool, an email archive and a notetaker, no single document can pull it together, and nobody has the hours to try. 

A meeting-summary tool does not close the gap either. It reads what was said in a conversation. The successor’s problem is everything that was already true before it. 

What continuity would actually require 

Set the tools aside for a moment and describe what would have to be true instead. 

The reading would have to come from the firm’s own records rather than from a transcript or a summary. The relationship record is where the history actually sits, and it is the only source that does not depend on someone having captured the right conversation at the time. 

It would have to happen without being asked. A successor opening a client record for the first time does not yet know what to ask about, which is exactly the moment a prompt-driven tool is least useful. The reading has to be on the page automatically, pre-written, and ready to read. 

And every claim in it would have to be checkable. Someone inheriting a relationship is in no position to judge whether an assertion about it is true, so each one has to trace back to the record it came from. 

Individually those three requirements are unremarkable. Together they describe something a document cannot be. 

Three readings that survive a handover 

Sentir puts sixteen specialist AI agents inside the system of record, each with a defined job and its own curated library of published reference material. They fall into three families: Operational Context reads the factual and operational state of a relationship; Domain Experts read a professional subject area against published authorities; and Analytical Perspectives read trajectory and tone. Three of them do the work a handover needs. 

Relationship Navigator maps who matters around the client. It reads the people and organizations connected to the record, their roles and relationship types, the composition of the household, and how that household connects to the trusts and entities around it. It reads the fiduciary and professional roles attached to those people, the agent under a power of attorney, the trustee, the guardian, the executor, against their established meanings rather than their labels. What comes back is the two or three connections that genuinely stand out, not a list of every contact on the file. 

Relationship Historian reads the arc. When the relationship began, the milestones and lifecycle transitions along the way, how the rhythm of contact shifted over the years, and which events turned out to be turning points. Every observation is anchored to a dated event rather than to how long ago something happened, so the reading holds whenever the successor opens it. 

Practice Manager reads the operational state. What is open and how long it has been open, which processes are in flight and how long each has held its current stage, which deliverables are due, and whether service reviews have kept to the expected cadence. This is the part of a handover that usually gets reconstructed by asking around. 

Two handovers 

The first call. A successor takes a call from a client who assumes the firm already knows their history. The exposure in that conversation is rarely a portfolio fact, which is written down and easy to find. It is a designation. The daughter who has been the point of contact for two years is also the agent under a power of attorney, which changes who should be on the call and what can be discussed without her. 

The book that changed hands mid-process. Three things were in flight when the relationship moved. One was a task with a due date, so it was visible. The other two were a process that had been holding at the same stage since spring and a review that had quietly slipped its cadence. Both were in the records. Neither was anywhere a successor would have thought to look. 

The Universal Timeline sits behind both. It carries five years of a relationship’s activity in three-month windows, on contacts, households, groups, divisions and the other record types it appears on, so a successor reads the sequence rather than assembling it. Manage Servicing Team handles the reassignment itself, and where a household spans several entities, Groups shows the structure as one picture rather than four separate records. 

What it will not do 

The constraints matter as much as the capability here, and in a handover they matter more, because a successor cannot yet tell a good inference from a bad one. 

It never infers meaning from an absence. A connection that is not in the records is not reported as a finding. Early versions of this kind of system fill their output with observations about what is missing, which tells a reader nothing they did not already know, so the agents are built to write about the records that exist. 

It orients rather than prescribes. Every specialist works with the same instruction: surface what is material and leave the decision to the advisor. No action plans, no prioritized next steps, no deadlines to act on. The agent does the assembly. The advisor brings the judgment. 

Each one stays in its lane. Where the shape of a household touches the estate mechanics behind an executor designation, Relationship Navigator names the connection in a clause and leaves the analysis to the specialist who owns it. 

And every factual claim traces to a source. Citations are matched to the underlying record by the platform rather than typed by the model and checked at the level of each paragraph and list item. A claim that cannot be grounded is not written. 

Where a successor starts 

The constraint in a handover was never expertise. A successor could always have read twenty years of records. There were never the hours before the first meeting, and there never will be. 

What changes is where they start. An advisor who opens an inherited record already oriented to the network around the client, the arc of the relationship and the work in flight has not been handed a summary of someone else’s knowledge. They have been handed a reading of the firm’s own records, which is a thing they can check, and which will still be there next quarter when the next relationship changes hands. 

At Practifi, that is the continuity we set out to build.