Summary
- File Check examines supporting records together, identifying inconsistencies between client circumstances, assessments and recommendations.
- BRI Wealth Management reports that many reviews have fallen from approximately 90 minutes to 20 minutes.
- Future development will extend checks to advisers before recommendations are executed, alongside risk-based review processes.
Edinburgh-based Aveni has introduced an artificial intelligence application that examines complete financial advice cases, aiming to reduce manual compliance work and make it practical for firms to review a greater proportion of their customer files.
File Check, launched on 16 November, compares information across advice documents and assesses cases against a firm’s own quality assurance framework. It is available to UK wealth management and financial advice businesses following development with selected industry partners.
Early customers have reported reductions of up to 78% in the time required for compliance reviews. BRI Wealth Management, which participated in the software’s development, says many checks that previously took approximately 90 minutes can now be completed in around 20 minutes.
The results indicate the potential to reduce routine document examination, although they do not establish average performance across all firms or independently demonstrate the system’s ability to identify every material compliance issue.
Advice reviews involve more than individual documents
A financial advice case can contain between 30 and 50 documents, according to Aveni, including suitability reports, fact finds, advice letters and records of client discussions.
Compliance teams must assess whether recommendations reflect the customer’s circumstances, whether the relevant information has been recorded and whether the supporting evidence satisfies regulatory and internal requirements.
Documents that appear complete when examined separately may contain contradictions when considered together. An initial record might identify one financial objective, for example, while a later recommendation relies on assumptions that are inconsistent with the client’s stated circumstances.
Identifying such discrepancies can require reviewers to compare information distributed across numerous files, often involving different formats and levels of detail.
File Check uses AI to examine the material collectively, looking for gaps between what the customer said, what the adviser assessed, what was recommended and what was documented. Its assessments are based on the firm’s own review questions, quality assurance framework and applicable regulatory principles.
This approach allows organisations to retain established compliance procedures rather than adopting an identical set of review criteria imposed on every customer.
However, the effectiveness of the checks will depend partly on the quality of the framework supplied. Applying an incomplete or outdated checklist consistently could make reviews faster without identifying all the issues that require attention.
The distinction between a missing record and unsuitable advice also remains important. A system may recognise that a particular explanation is absent, while determining whether the underlying recommendation was appropriate requires further information and professional judgement.
Early customer results indicate faster reviews
BRI Wealth Management has worked with Aveni on the product since its early development and reports that many compliance checks now require approximately 20 minutes rather than an hour and a half.
The reduction is consistent with Aveni’s reported saving of up to 78%, but it relates to particular cases and circumstances. Review times may vary considerably according to document quality, advice complexity and the amount of further investigation required.
A comparison of processing times alone also provides limited evidence about accuracy. Organisations adopting automated compliance tools need to assess whether the system identifies material issues reliably and how frequently reviewers must investigate findings that turn out to be immaterial.
These questions become more important if automation enables firms to examine substantially more files. Many advice businesses use risk-based sampling because reviewing every case manually would require extensive resources.
Reducing the effort involved in initial checks could make broader coverage commercially practical, while allowing experienced compliance staff to concentrate on cases requiring detailed examination.
Nevertheless, reviewing every file does not guarantee that every concern will be identified. Some failures may originate in information that was never recorded, inaccurate assumptions supplied by a customer or circumstances that an automated system cannot interpret reliably.
Firms therefore need procedures for handling uncertain results, escalating concerns and documenting the reasons for consequential decisions.
Checking cases before advice is executed
File Check currently supports compliance teams and employees responsible for testing customer outcomes. Aveni intends to extend the technology into the work of advisers and paraplanners so that potential problems can be identified before recommendations are executed.
The planned functionality would allow staff to examine case records during preparation, correcting incomplete information or inconsistent documentation before the customer acts on the advice.
Aveni expects these checks to become accessible through File Check and its existing Aveni Assist workspace, with additional integrations into customer relationship management systems also planned.
These capabilities remain part of the development roadmap and should not be confused with the functionality already available to compliance reviewers.
The proposed operating model combines checks before execution, risk-based triage and continuing assurance after advice has been provided.
Under that approach, firms could give every case an initial assessment while directing more extensive human review towards higher-risk circumstances. Its effectiveness would depend on the organisation’s risk criteria and the reliability with which cases are classified.
Earlier checks could reduce the need to revisit completed recommendations or contact clients to correct records. They could also help identify information that requires clarification before an adviser reaches a final conclusion.
However, placing automated checks within advisers’ daily workflows creates a risk of overreliance on apparently satisfactory results. Staff must understand when additional investigation is necessary and should not treat an automated assessment as a substitute for their professional responsibilities.
Specialist models support financial terminology
File Check uses the Aveni Intelligence Engine, which incorporates models developed for financial services and technology already deployed through Aveni Detect.
The system includes FinLLM, Aveni’s specialised language model for regulated financial environments.
Financial advice documents can contain terminology and relationships that differ from ordinary business correspondence. Interpreting suitability requires understanding how a client’s objectives, knowledge, circumstances and tolerance for risk relate to the recommendation being made.
A specialist model may improve the handling of this information, but its suitability must still be demonstrated through testing across representative advice cases. The existence of a sector-specific training approach does not independently establish superior accuracy.
Information security also requires attention because financial advice files contain sensitive personal and financial records. Firms must examine access permissions, processing arrangements, retention periods and the evidence available when automated assessments are challenged.
The Financial Conduct Authority’s Consumer Duty requires authorised firms to act to deliver good outcomes for retail customers. AI-assisted review may support compliance monitoring and evidence gathering, but responsibility for the advice and its consequences remains with the regulated business.
File Check represents a practical application of AI within a document-intensive professional process, where reducing review time could improve the coverage of existing controls.
Its early customer results provide evidence of potential efficiency gains, while wider deployment will establish whether those improvements can be maintained alongside consistent identification of material issues across different firms, advice types and customer circumstances.












