What problem does it solve?
After every client meeting an advisor has to write a file note, record what the client said about goals, risk and circumstances, list what was agreed, send a follow up and update the CRM. With the next meeting waiting, this work is easy to postpone, and a note written days later can miss what the client actually said, which weakens the firm's record when a recommendation is later questioned.
The work is also expensive: it takes senior, client facing time, or an assistant who sat in the meeting. And when notes are short, the CRM, which should hold the richest view of the client, holds little of the conversation, so the next meeting starts from memory.
How does it work?
- Consent first. The advisor tells the client the meeting will be transcribed by an AI notetaker and records the client's consent; without it, nothing is recorded.
- Capture. The notetaker joins the video call or the phone line and produces a transcript with speaker separation.
- Draft the file note. A summary is generated in the firm's template: attendees, topics, client goals and circumstances mentioned, decisions, action items with owners, and anything the client asked to be checked.
- Advisor review. The advisor corrects and approves the note and the draft follow up email. Nothing is filed or sent without that approval.
- Write back. The approved note, tasks and key facts are written into the CRM and the transcript is retained under the firm's record keeping policy.
- Audience
- Employee facing
- Autonomy
- Copilot
- Adoption
- Mainstream
- Channels
- Microsoft Teams, Phone and voice, Internal tools, Email
What is it worth?
Benchmarks are computed from the public deployments below: one data point per organization per KPI, with who made each claim.
| KPI | Median | Reported range | Data points | Claimed by |
|---|---|---|---|---|
| Productivity gain | Too few to pool | 15% | 1 | 1 vendor |
| Time saved per task | Too few to pool | about 30 minutes | 1 | 1 vendor |
Value drivers: Employee productivity, Compliance quality, Customer experience.
Indicative value
A wealth manager with 500 client facing advisors
USD 1.8 million to USD 13.8 million
Value of advisor time released from meeting write up per year
How this is calculated
Formula: advisors * meetingsPerWeek * minutesSaved / 60 * weeks * hourlyCost. The low scenario uses every low input, the high scenario every high input.
| Input | Low | High | Basis |
|---|---|---|---|
| Client facing advisors advisors, advisors | 500 | 500 | The reference firm. |
| Client meetings per advisor per week meetingsPerWeek, meetings per advisor per week | 3 | 6 | Editorial assumption, replace with your own CRM activity data. |
| Minutes of write up saved per meeting minutesSaved, minutes per meeting | 20 | 40 | Brackets the one reported saving on this page (UniSuper, about 30 minutes per interaction) and sits below the 45 minutes per meeting a Quilter Cheviot investment manager assumes in the firm's estimate; the low end allows for review time. Replace with your own pilot data. |
| Working weeks per year weeks, weeks per year | 46 | 46 | Editorial assumption. |
| Fully loaded advisor cost per hour hourlyCost, USD per hour | 80 | 150 | Editorial assumption, replace with your own fully loaded cost. |
What it leaves out: Counts released time only. It leaves out licence and transcription costs, the time to review each note, and the harder to price benefit of a more complete record in complaints and audits.
Who already uses it?
6 public deployments, strongest evidence first. Grades: A regulator or audit, B the organization itself, C vendor case study, D anonymous or estimate.
Bank of America
United States · Wealth and asset management · 2026
Merrill Wealth Management and Bank of America Private Bank rolled out an AI meeting solution at full scale in March 2026. It consolidates client relationship insights and recent activity into meeting preparation material, takes notes in virtual meetings with client consent, and turns the decisions into a summary, tasks and documentation afterwards. The bank says the capability can save advisors up to four hours per meeting; it presents this as potential, not as a measured result, so it is not recorded as a metric here.
No outcome disclosed.
Morgan Stanley
United States · Wealth and asset management · 2024
Morgan Stanley Wealth Management launched AI @ Morgan Stanley Debrief in June 2024. With client consent, the tool takes notes in client meetings, surfaces action items, summarizes the key points, drafts a follow up email for the advisor to edit and send at their discretion, and saves a note into Salesforce. The release quotes advisors on the time saved on note taking (one cites about half an hour per meeting) but gives no firm wide measurement.
No outcome disclosed.
Quilter
United Kingdom · Wealth and asset management · 2025
Quilter, a UK wealth manager, rolled out Microsoft 365 Copilot and names meetings and transcriptions as its biggest use case. Microsoft reports that Quilter estimates Copilot will save more than 13,000 hours per month of post call admin time; an investment manager at Quilter Cheviot builds that estimate from an assumed 45 minutes saved per client meeting across 174 investment managers doing about 100 meetings each. Both figures are projections, not measured savings, so neither is recorded as a metric. Quilter also tested turning a portfolio manager interview transcript into an investment commentary: about 15 minutes of prompting and half an hour of editing instead of a few days, which it describes as a one off test.
No outcome disclosed.
SEB
Sweden · Banking · 2025
SEB, a Nordic corporate bank, worked with Bain & Company to build an AI agent on Google Cloud for its wealth management division. The agent suggests responses during conversations with customers and generates call summaries afterwards. Google Cloud reports a 15% efficiency gain.
- Productivity gain: 15%
"The agent, built with Google Cloud, enhances end-customer conversations with suggested responses and generates call summaries, helping to increase efficiency by 15%."
Claimed by: vendor
Commerzbank
Germany · Banking · 2024
Commerzbank implemented an AI agent on Gemini 1.5 Pro that automates the documentation of client calls, a manual task for its financial advisors. Google Cloud reports a significant reduction in processing time, which lets advisors spend more time with clients, but gives no figure.
No outcome disclosed.
UniSuper
Australia · Wealth and asset management · 2024
UniSuper, an Australian superannuation fund, uses Microsoft 365 Copilot to produce file notes that summarise the key details of each adviser conversation with a member held on Microsoft Teams. Microsoft reports advisers save roughly 30 minutes per client interaction; the fund also says the notes give it better visibility of interaction quality. The annual hours and extra members advised are projections and are not recorded.
- Time saved per task: about 30 minutes, per client interaction
"Advisors are saving roughly 30 minutes per client interaction on Microsoft Teams by using automated, bespoke file notes that summarise key details of each conversation."
Claimed by: vendor
How do you implement it?
A model agnostic playbook: what to prepare, the order to build in, and what goes wrong.
Data you need
- An agreed file note template per meeting type (review, new advice, service)
- Consent wording and a place to record the client's consent
- Retention and deletion rules for audio, transcripts and notes
- CRM field mapping for notes, tasks and client facts
Systems to integrate
- Video meeting and telephony platforms
- Speech to text with speaker separation
- CRM such as Salesforce or Microsoft Dynamics 365
- Email for the draft follow up
- Records management or archive for retained transcripts
Complexity: Medium
Transcription and summarization are mature. The work is in consent capture, the note template, retention rules for recordings and transcripts, and a reliable write back into the CRM.
- 1
Agree the note standard with compliance
Define what a good file note contains for each meeting type and which statements must be captured verbatim (for example a client's stated risk appetite or a refusal of advice).
- 2
Design consent and retention
Write the consent script, record consent in the CRM, and decide how long audio and transcripts are kept and where. Some clients will decline; the process must work without AI.
- 3
Pilot with a small group of advisors
Measure time to a finished note, edit rate and advisor satisfaction per meeting type, and collect notes that went wrong to improve the template and prompts.
- 4
Automate the write back
Once notes are reliable, write the approved note, tasks and structured facts into the CRM through its API rather than copy and paste.
- 5
Sample and supervise
Supervisors review a sample of notes against transcripts each month, with extra attention to meetings where advice was given.
Guardrails
- No recording without recorded client consent, with an easy opt out
- Advisor approval before any note is filed or any message is sent
- Notes limited to what was said; no inferred emotions, health conditions or vulnerability labels without a human decision
- PII masking and access control on transcripts, with retention limits
- Every AI generated note labelled as such in the CRM
KPIs to instrument
- Median minutes from meeting end to approved note
- Share of meetings with a complete note within 24 hours
- Advisor edit rate per note section
- Consent rate and opt outs
- Supervisor sample findings per month
Human in the loop
The advisor reviews, corrects and approves every note and follow up; supervisors sample notes against transcripts. Suspected vulnerability or complaints mentioned in a meeting go to a human process, not to an automated flag alone.
Common failure modes
- Rubber stamped notes
- Advisors approve drafts without reading them, so errors enter the record. Track time spent reviewing and sample notes against transcripts.
- Missing or misattributed statements
- Speaker separation fails on a phone line and a client's words are attributed to the advisor. Test on real audio and flag low confidence passages.
- Consent gaps
- Recording starts before consent is captured, or consent is not stored. Make consent a hard gate in the workflow.
- Over retention
- Audio and transcripts are kept indefinitely by default. Apply retention rules from day one.
What are the risks and rules?
EU AI Act
Minimal risk
Transcribing and summarizing meetings for an employee is not a use listed in Annex III, and the advisor reviews every note before it is filed or sent. The tier would change if the tool inferred emotions: emotion recognition is high risk under Annex III point 1(c), and inferring the emotions of employees at work is prohibited under Article 5(1)(f). Both stay out of scope.
Rules that apply
Guidance
- ESMA public statement on the use of AI in the provision of retail investment services (European Securities and Markets Authority, Europe). Expects investment firms to keep comprehensive records on their use of AI, and applies MiFID II conduct and record keeping duties to AI supported advice.
- Principles to Promote Fairness, Ethics, Accountability and Transparency (FEAT) (Monetary Authority of Singapore, Asia Pacific). Transparency and accountability principles for data driven tools used in client relationships.
Controls to put in place
- Consent capture and storage for every recorded meeting
- Retention schedule for audio, transcripts and notes aligned with record keeping rules
- Labelling of AI drafted notes and an audit trail of advisor edits
- Monthly supervisory sampling of notes against transcripts
- Inventory entry with an accountable owner and a documented note template
Frequently asked questions
- How much time does an AI notetaker save an advisor?
- The one reported saving is about half an hour: Microsoft reports UniSuper advisers save roughly 30 minutes per client interaction. Quilter's estimate of more than 13,000 hours a month assumes 45 minutes saved per meeting, which is a projection rather than a measured result. Bank of America says the capability can save advisors up to four hours per meeting, which it presents as potential rather than a measured result.
- Do we need client consent to record and transcribe meetings?
- Morgan Stanley and Merrill both run their notetakers with client consent. Whether consent is legally required depends on the jurisdiction and the lawful basis you rely on, so agree it with legal and compliance, and make sure the process still works when a client says no.
- Can the AI note replace the advisor's own record?
- No. The note is a draft; the advisor approves it and remains responsible for its accuracy. Supervisors should sample notes against transcripts, especially for meetings where advice was given.
How to cite this page
Blits.ai AI Use Case Library, "AI meeting notes and CRM update for wealth advisors", last verified 27 September 2026, https://www.blits.ai/ai-use-cases/client-meeting-notes-and-crm-update. Licensed under CC BY 4.0. Method: how we verify use cases.
Changelog
- 27 September 2026: First published