Wealth · Suitability records
WorkflowWealth managementChecking every suitability report before the file review does
File reviews in wealth management usually sample a few advice files a quarter. Here is how one workflow can check every suitability report for missing elements, and why the judgement on suitability stays with a person.
Valnora & Elric team5 min read
Questions in this note
These are working notes on a narrow problem in wealth management: making sure every suitability report contains what your own file standard says it must, before a file review or a regulator finds out it does not.
Why is a sample-based file review not enough?
Most firms check advice quality by sampling a few files per adviser each quarter. That is the right way to judge advice, because judging whether a recommendation suited a client takes an experienced reviewer and time.
It is a poor way to catch missing paperwork. A report without a capacity for loss assessment or a costs figure is a record-keeping gap whatever the quality of the advice, and a sample will catch only some of them. The record requirements are explicit in most regimes. In the EU, MiFID II and its delegated rules require firms to assess suitability and set out what the suitability report must cover. In the UK, COBS 9A carries the equivalent rules.
What exactly does the workflow check?
Only the elements your file standard lists. A typical standard includes:
- the client’s objectives and time horizon
- risk tolerance, and how it links to the recommendation
- capacity for loss
- knowledge and experience of the products recommended
- the reasons the recommendation meets the client’s objectives
- costs and charges
- whether the client needs periodic review
Each element gets one of three results: present, missing or unclear. Every result must quote the sentence it relied on, or say that no sentence was found. A reviewer can then confirm the result in seconds instead of reading the full report.
How is the workflow built?
One workflow from new report to reviewer queue
Workflow · Suitability record check
- 01Daily scheduleRuns each morning over reports added to the advice folder the day before.
- 02Read the reportThe File block loads each report. AWS Textract reads scanned or signed copies.
- 03File standardYour suitability file standard and house policy, searched in a Knowledge Base.
- 04Check each elementPresent, missing or unclear for each element, with the sentence that shows it.
- 05Check registerOne row per report with each result, the adviser and the client reference.
- 06Reviewer queueFiles with gaps go to a file reviewer, who decides what the adviser must fix.
It is one workflow in the workflow builder, run daily on a schedule.
- Pick up new reports. A File block loads reports added to the advice folder since the last run. Scanned or signed copies go through the AWS Textract block first.
- Load the standard. A Knowledge block searches your file standard and house policy in the Knowledge Base, so the check uses your wording rather than the model’s idea of a good report.
- Check each element. An Agent block returns a fixed set of fields, one per element, each with a result and a quote.
- Record it. A Table block writes one row per report to a check register in Tables, with the adviser and client reference.
- Route gaps. A Condition block sends reports with any missing or unclear result to the reviewer queue. Complete reports are simply logged.
What does a result look like?
What the solution produces
Suitability report check (illustrative)
Checklist- Client objectives recorded: present. "Grow the portfolio for retirement in about 15 years."
- Risk tolerance recorded and linked to the recommendation: present. Risk profile 4 of 7.
- Capacity for loss assessed: unclear. Mentions emergency savings but no view on the impact of a fall in value.
- Knowledge and experience recorded: present.
- Reasons the recommendation meets the objectives: present.
- Costs and charges disclosed: missing. No ongoing charge figure in the report.
- Need for periodic review stated: present. Annual review offered.
The check is deliberately literal. "Unclear" is the most useful result: it usually means the adviser wrote something, but not enough for a reader to follow the reasoning. Those are the cases where a short note to the adviser fixes the file.
Does the workflow judge whether the advice was suitable?
No, and it should not. Whether a recommendation was suitable depends on the client’s full circumstances, the alternatives the adviser considered and conversations that may not be in the report. That judgement stays with your file reviewers and, where needed, compliance.
The workflow’s instructions say this directly, and its output has no field for "suitable" or "unsuitable". What it gives reviewers is a complete list of documentation gaps, so their sampled reviews can focus on the advice itself.
How do reviewers and compliance see the results?
The check register is a normal workspace table. File reviewers work from the queue, and compliance can filter by adviser, team or element to spot patterns, such as one team that never records capacity for loss. Row-level access policies can limit each team leader to their own advisers.
Every run is traced block by block in Logs, so for any report you can show which version of the standard was searched and what the check returned.
What the change could be worth
An advice firm issuing 400 suitability reports a month and file-reviewing a 10% sample.
| Assumption | Value |
|---|---|
| Suitability reports issued per month | 400 reports / month |
| Share of reports file-reviewed today | 10% |
| Share of reports with a documentation gap | 8% |
| Share of missed gaps later needing remediation | 50% |
| Cost to fix a gap when the report is written | 40 EUR |
| Cost to remediate a gap found later | 400 EUR |
- more record gaps caught a month: Suitability reports issued per month × Share of reports with a documentation gap × (1 − Share of reports file-reviewed today)
- remediation cost avoided a year: Suitability reports issued per month × Share of reports with a documentation gap × (1 − Share of reports file-reviewed today) × Share of missed gaps later needing remediation × (Cost to remediate a gap found later − Cost to fix a gap when the report is written) × 12
Modelled from the assumptions above to show how the outcome is calculated; not a customer result. Your results depend on your volumes and process.
Controls and governance
Run logs
Every workflow run records each block’s inputs, outputs and errors, so a reviewer can trace how a result was reached.
Deployed versions
Workflows run from a deployed version, so a change is reviewed before it reaches production runs.
Row-level access
Row-level policies narrow what each person sees in tables, dashboards, exports and scheduled reports.
Roles and permission groups
Admin, write and read roles per workspace, plus permission groups that limit which models, blocks and integrations a team can use.
PII detection
A Guardrails block detects personal data and blocks or masks it before text reaches the next step.
Self-hosting
Run Agent Serve in your own cloud or data centre, next to the systems it reads.
Bring your own keys
Use your organisation’s own model provider keys at organisation or workspace level.
More on how Agent Serve handles access, data and audit on the security page.
Where should a firm start?
Take last quarter’s reports for one adviser team and run the workflow over them. Compare its results with your file reviewers’ notes for the files they sampled. Where they disagree, tighten the wording of the standard rather than the prompt; ambiguity there usually confuses people too.
For the conversation side of the same advice process, see disclosure and consent checks on calls, and for preparing advisers before meetings, see relationship manager meeting prep.
Questions teams ask
Can the workflow check reports written before the standard changed?
Yes, if you keep each version of the file standard in the Knowledge Base with its start date. Ask the Agent block to apply the version in force on the report date.
Does it work on fact-finds as well as reports?
It can, by adding the fact-find as a second document and a check that the report reflects what the fact-find recorded. Start with reports alone, then add this once the first checks are reliable.
What you need
- Access to the folder or document system where final suitability reports are saved
- Your suitability file standard, written as a list of elements a report must contain
- Your house policy and any adviser guidance, in a Knowledge Base
- An adviser and client reference on each report, so results can be grouped
- Named file reviewers who own the reviewer queue
- Agreement with compliance on what a missing or unclear result triggers