Collections Case

Collections cases exist only as verbal handoffs between loan officers and collections staff. When a borrower misses payments, the branch manager mentions it in a morning meeting or sends a quick email to the workout team. There is no written case record — the collector relies on memory, sticky notes, or personal notebook entries to track which delinquent borrowers they have contacted and what was discussed. If the assigned collector is absent, no one knows which accounts are in active workout, what promises to pay were made, or whether any skip-tracing efforts are underway. Charge-off decisions are made based on the collector's verbal summary to a supervisor, with no documented trail of collection activity. Regulatory examiners asking for collections activity logs receive apologetic explanations that records were not kept. When borrowers dispute the collection timeline or claim they were never contacted, the institution has no evidence to support or refute those claims. Recovery rates are unknown because no one tracks the progression from delinquency through resolution.

None — AI cannot assist with collections prioritization, contact scheduling, or recovery prediction because no machine-readable collections case record exists.

Collections cases are tracked in personal spreadsheets or shared Word documents maintained by individual collectors. Each collector has their own format — one might track by borrower name and phone number, another by loan number and last contact date. When a supervisor asks for the total number of active workout cases, collectors manually count their rows and report back. Promise-to-pay agreements are noted in free-text comments like 'borrower said they will pay $500 by Friday' with no standardized fields for amount, date, or follow-up action. Skip-tracing results appear as notes such as 'tried new number, no answer.' Charge-off recommendations are written in emails to the credit officer with varying levels of detail. Some collectors maintain excellent personal records; others have gaps spanning weeks. There is no way to see the full collections pipeline or assess whether the team is following consistent contact strategies across all delinquent accounts. Charge-off timing is inconsistent because there is no documented standard for when a delinquent loan should be written off versus continued in workout. Different collectors apply different thresholds based on personal experience, and the institution cannot demonstrate a consistent loss recognition practice to examiners. The lack of formality means that similar delinquent accounts receive different treatment depending on which collector is assigned.

AI could scan collector spreadsheets for keywords like 'promise to pay' or 'charge-off,' but cannot reliably extract structured payment commitment data or calculate portfolio-level delinquency trends from inconsistent free-text formats.

Collections cases follow a standardized template with consistent sections: borrower identification, loan details, delinquency classification, contact history log, promise-to-pay commitments, and case status. Every collector uses the same form, and required fields are enforced before a case can be updated. The template includes dropdown menus for contact outcome codes such as 'right-party contact,' 'left message,' 'wrong number,' and 'promise to pay.' However, the case record lives as a document in a shared network drive or a basic tracker rather than a purpose-built collections system. Supervisors can review individual case files but cannot easily query across all cases to find, for example, every account with a broken promise-to-pay in the last thirty days. Charge-off packages are assembled manually by copying information from the case template into a separate recommendation form. The collections framework provides clear escalation paths from early-stage delinquency through late-stage workout to charge-off or recovery. Each stage has defined entry criteria, required documentation, and expected timelines. Collectors can quickly determine where any account stands in the lifecycle and what actions are required next. Management reporting draws directly from the structured case data to produce regulatory call report figures and internal loss forecasting.

AI can search and retrieve collections case documents by borrower name or delinquency bucket, but cannot programmatically aggregate promise-to-pay fulfillment rates or identify collectors with the highest right-party contact ratios without manual data extraction.

Collections cases are managed in a dedicated collections system with discrete fields for borrower demographics, loan details, delinquency aging bucket, contact attempt log with timestamps, contact outcome codes, promise-to-pay records with amount and target date, skip-tracing results, and workout strategy assignment. The system enforces mandatory documentation before status transitions — a collector cannot mark a case as 'resolved' without entering a resolution code and payment confirmation. Each case carries a unique identifier linked to the originating loan record. Supervisors run daily reports showing cases by aging bucket, contact attempts per case, promise-to-pay pipeline, and charge-off queue. The system timestamps every action, creating an audit trail that satisfies examiner requests for collections activity documentation. The collections framework provides clear escalation paths from early-stage delinquency through late-stage workout to charge-off or recovery. Each stage has defined entry criteria, required documentation, and expected timelines. Collectors can quickly determine where any account stands in the lifecycle and what actions are required next. Management reporting draws directly from the structured case data to produce regulatory call report figures and internal loss forecasting.

AI can prioritize the collections queue by scoring accounts based on delinquency severity, payment history patterns, and contact success rates. Automated dialer integration can schedule outbound calls based on optimal contact time analysis. Promise-to-pay tracking with automated reminder generation is feasible.

Collections cases are schema-driven entities with explicit relationships to the originating loan, borrower profile, collateral valuation records, prior workout history, bankruptcy filings, and regulatory compliance flags. Each case carries a machine-readable workout strategy tree showing which interventions were attempted, in what sequence, and with what outcomes. The system understands that a 90-day delinquent commercial real estate loan with declining collateral value requires different treatment than a 30-day delinquent consumer auto loan. An AI agent can query 'show all cases where the borrower has broken two or more promise-to-pay commitments and the collateral LTV now exceeds 100 percent' and receive a structured result set with recommended next actions. Charge-off analysis includes net present value calculations of recovery scenarios. The institution can demonstrate to regulators and auditors that every delinquent account follows a consistent, policy-driven lifecycle from initial delinquency notification through final resolution. Cross-portfolio analysis reveals which workout strategies produce the best recovery rates for different borrower segments, property types, and delinquency causes. The formalized framework has reduced average time-to-resolution and improved net recovery rates measurably.

AI can generate dynamic workout strategy recommendations based on borrower behavior patterns, collateral trends, and recovery probability models. Automated settlement offer generation with authority-level routing and regulatory compliance checks is possible.

Collections case records are living entities that self-assemble from loan servicing feeds, payment processing systems, credit bureau updates, collateral monitoring services, and regulatory rule engines. When a loan crosses a delinquency threshold, the system automatically instantiates a case record, populates it with borrower contact history from across all relationships, pulls current collateral valuations, checks for bankruptcy filings and regulatory holds, and assigns an initial workout strategy based on the institution's current loss mitigation policies. The case schema adapts dynamically — when new regulations modify collections contact frequency limits or when the institution adjusts its charge-off timing policies, case validation rules and workflow triggers update automatically. The collections case is a real-time reflection of the borrower's workout status against current institutional policy and market conditions, continuously enriched by external data feeds. The institution can demonstrate to regulators and auditors that every delinquent account follows a consistent, policy-driven lifecycle from initial delinquency notification through final resolution. Cross-portfolio analysis reveals which workout strategies produce the best recovery rates for different borrower segments, property types, and delinquency causes. The formalized framework has reduced average time-to-resolution and improved net recovery rates measurably.

Fully autonomous early-stage collections management is possible. AI can execute contact strategies, negotiate payment arrangements within delegated authority, escalate complex cases with full context packages, and optimize portfolio-level recovery rates through continuous strategy refinement without human intervention for standard delinquency scenarios.

Ceiling of the CMC framework for this dimension.