AI-Driven Regulatory Reporting: Automated Preparation of Y-9C Schedules with ML/AI

One of our clients, a financial investment firm, faced challenges preparing Y-9C reports manually. Preparing Y-9C schedules involves gathering financial data from multiple internal systems, mapping it correctly, and ensuring alignment with regulatory instructions. This work is typically performed by regulatory reporting teams, who rely on manual processes to extract, organize, and validate figures. Given the short timelines, high data volume, and repetitive nature of the task, the process is resource-intensive and exposed to operational risk.

Challenge

• Analysts navigate multiple systems to retrieve and format data.
• Mapping rules and adjustments are often documented informally or spread across legacy files.
• Even minor changes in data or instructions require complete revalidation.
• Manual execution increases the risk of inconsistencies or last-minute errors.

As a result, analysts often spend more time assembling the numbers than analyzing or explaining them.

Delivery

We developed an end-to-end automation pipeline that combines business rules and machine learning to prepare draft schedules:

• Rule-Based Layer: Known mappings, thresholds, and recurring adjustments were encoded as Python logic to automate predictable, stable cases.
• ML Layer: A supervised learning model was trained using historical submissions to identify how data points are typically mapped across reporting lines. This enables the system to handle exceptions and adapt to evolving patterns in the data.
• Traceability Layer: Every value generated by the system includes a detailed explanation. Whether produced by rule or model, the system records:
  • Prediction source (rule-based or ML)
  • Confidence score
  • Key drivers (feature importance using SHAP or similar method)
  • Historical mapping for the same field in previous quarters
  • Peer comparisons from similar data sets
  • Full audit trail of the data source and transformation steps

This framework allows analysts to understand how each value was derived.

Value

Automating Y-9C schedule preparation does not eliminate the analyst’s role-it enhances it:

• From manual assembly to strategic review: Analysts focus on validating pre-filled schedules rather than starting from scratch.
• More time for impact analysis: Teams can concentrate on interpreting changes and preparing explanations for management and regulators.
• Reduction in manual errors: Logic is consistently applied, and the ML model improves with each cycle.
• Full traceability: Every number is backed by transparent logic and a clear audit trail, supporting regulatory confidence.
• Scalable design: The system can quickly adjust to new reporting requirements, data changes, or structural updates.

The use of AI in this workflow transforms the analyst’s role from preparer to reviewer. By automating repetitive tasks and surfacing clear justifications for each number, analysts can focus on reviewing results, understanding changes, and ensuring consistency.