Profile risk analysis
Risk summaries score borrowers 0–100 with grade (A–D), risk level, factor breakdown, and actionable recommendations cached for 24 hours.
AI Review Layer
A full AI banking agent, grounded in your borrower data.
The AI does the work of a banking agent: it reads documents, builds client profiles, prepares credit memos and facility letters, schedules meetings, drafts communications, answers staff questions and monitors accounts, with every action logged and reviewable.
Overview
AI prepares the decision. Your credit officers approve it.
The AI Review Layer sits inside the Management Portal and works as a banking agent: a conversational assistant, OCR and document extraction, client profile building, credit memo and facility letter preparation, draft communications and account monitoring. Your credit officers approve every decision, with four-eyes, approval matrices and a full audit trail.
On each profile, AI Risk Analysis generates a scored summary with grade, risk factors, and recommendations. In the assistant, staff ask natural-language questions and the model invokes tools to look up borrowers, rank risk, list overdue notices, and — for admins — propose and apply governed edits with a two-step confirmation flow.
Summarize, extract, govern
Three jobs, each approved by a person.
Email and document intake
Inbound emails and attachments are parsed into structured applications, and documents are read by OCR with each figure linked to its page and line. Low-confidence extractions route to a review queue for staff correction before approval.
Prompt and edit controls
Admin prompt configuration aligns summary format and emphasis. Copilot edits require explicit user confirmation and are audit-logged with before/after previews.
Capabilities
What the AI Review Layer delivers to staff.
Summaries, extraction, and conversational access — all scoped to the borrower's real record and your team's permissions.
- 01
AI risk summary
On-demand risk analysis on any borrower profile. Score breakdown chart, listed risk factors and strengths, with the rules-based score, risk band, and recommendation kept separate from AI narrative support.
- 02
Copilot assistant
A chat assistant with eight tools. Staff ask questions in plain language; the model looks up borrowers, aggregates stats, and lists overdue notices.
- 03
Auto-resolved context
HKID, application number, loan number, mobile, and name patterns in chat messages auto-load borrower context — so common queries answer in one round-trip without manual lookup.
- 04
Email AI extraction
Connected mailboxes fetch unread messages, parse body and attachments, skip duplicates by message ID, and create structured lead records with a confidence score.
- 05
Inbound lead review
Low-confidence parses wait for review with an alert to admins. Staff correct fields, approve into live records, or reject with the audit context preserved.
- 06
Prompt administration
Admin prompt controls align summary format, risk emphasis, and extraction rules, with history available for cost and quality monitoring.
Credit review
From profile open to risk-informed decision
The AI prepares the summary and the recommendation; credit officers review and approve.
- 01
Open the profile
Open a borrower profile and trigger AI Risk Analysis on the summary tab.
- 02
Read it against the record
Review score, grade, risk factors, and recommendations alongside raw loan history.
- 03
Ask follow-up questions
Ask the Copilot follow-up questions — "show cross-company exposure" or "list overdue notices for this borrower."
- 04
Note the decision
Compare AI output against manual review of application info, DSR and credit reference, then note findings in case notes for committee or audit trail.
Intake extraction
From inbound email to structured lead
Operations teams triage AI-parsed leads before they become live borrower records.
- 01
Fetch
Scheduled or manual email sync fetches unread messages from connected mailboxes.
- 02
Extract
AI extracts name, HKID, mobile, loan amount, and purpose into lead fields.
- 03
Route
High-confidence parses can auto-route; low-confidence items queue for staff review.
- 04
Approve
Staff correct extraction errors and approve the lead into a borrower record. Every sync run reports how many items it processed, saved and failed on.
Copilot tools and surfaces
Eight tools. Governed edits. Full audit trail.
Read tools are open to all signed-in staff, scoped by company; write tools are for admins only. Every chat query and applied edit is logged.
Look up a borrower
Search by HKID, name, application number, loan number or mobile, with up to ten matches.
Full profile
The complete merged profile with the rules-based risk computation, once one match is identified.
Top risk borrowers
Rank borrowers by risk score, active loans or total loans within company scope.
Portfolio statistics
Counts by company, status, risk level or grade, with each borrower counted once.
Filtered lists
Paginated lists of blacklisted, high-active-loan or company-specific cohorts.
Overdue notices
Live overdue and upcoming notices, filtered by mode, company or borrower.
Propose an edit
Admin only: the borrower, the field, the before and after values, and ten minutes to confirm.
Apply an edit
Applies a confirmed proposal only after explicit approval, then writes the change to the audit trail.
Scope
A banking agent with explicit review and audit controls.
Who uses it
Credit reviewers, managers, intake admins and operations users who need database-backed portfolio answers, risk commentary or draft communications.
Included
Copilot tools, borrower lookup, portfolio aggregation, overdue-notice answers, document extraction, credit memo and facility letter preparation, risk commentary, collections drafts, meeting scheduling, email-intake assistance, early-warning monitoring and prompt controls.
Controls
Scoped read tools, admin-only write tools, propose and apply confirmation, protected source-system fields, prompt administration, and audit logs.
Boundaries
Credit approval stays with your credit officers: the agent prepares and recommends, and works within staff authority and the approval matrix. Borrower-facing assistants and custom agent skills are built to your bank as part of the implementation.
Demo Bring sample review questions, the risk summary format you want, your prompt rules, who may edit records, and how extracted email leads should be handed to staff.
Align AI review with your credit process.
Request a demo to review risk summary formats, email extraction rules, Copilot tool access, and the admin prompt controls that keep output consistent with your team.
Book a personalised demo