The Business Need

Our client, a prominent Edinburgh-based financial institution, faced challenges processing regulatory documents required for Know Your Customer (KYC) and Foreign Portfolio Investor (FPI) compliance. They aimed to improve both compliance and operational efficiency. To do this, they needed a solution to streamline the extraction, classification, and validation of key documents, including KYC forms, Ultimate Beneficial Owner (UBO) declarations, and Securities and Exchange Board of India (SEBI) registration certificates. This called for an intelligent, automated workflow that could handle document processing, metadata extraction, and real-time verification all in one place.

Challenges

Complex document structures

Handling complex multi-page documents with data scattered across different sections made accurate and timely data extraction a significant challenge.

Authentication and consent delays

Automating logins to KRA platforms involved OTP-based two-factor authentication. This led to delays due to verification lags and managing client consent. OTP issues on mobile further slowed onboarding, negatively impacting the user experience and compromising real-time access to client records.

Stringent regulatory requirements

Ensuring compliance with strict regulatory standards, especially when dealing with sensitive customer data, was challenging. It required robust systems for data validation, secure access, and maintaining audit trails.

Manual data management

Relying on manual data entry and verification was time-consuming and prone to errors. This increased the risk of non-compliance and reduced operational efficiency.

Data integration complexity

Integrating KYC data from multiple sources and aligning it with existing systems posed significant difficulties. The lack of smooth data flow hindered efficient decision-making.

XDAS Approach

To meet the client’s KYC processing automation needs, the XDAS team developed a structured, AI-driven workflow to streamline the overall process.

Automated document retrieval

We deployed bots to automate OTP-based two-factor to get the KYC Registration Agency platforms. The system handled OTP-based authentication and fetched client records. Consent workflows were managed securely, ensuring seamless access.

Seamless document preprocessing

The PDF Merger bot combined multiple documents related to each PAN into a single structured file. Non-readable PDFs were converted into machine-readable format using OCR. Following this, the PDF Chunker segmented the document page by page to streamline further processing.

Context-aware intelligent classification

LLM-Powered bots identified the category of each page based on its content rather than just keywords. The system pinpointed the latest authorized signatory pages and ensured documents were accurately split by category, allowing for focused processing.

Metadata extraction and validation

Using LLMs, key fields such as SEBI registration numbers, fund categories, and UBO details were extracted. These were validated against the FPI portal to ensure compliance. The system prioritized the most recent document versions, especially for authorized signatory lists, to maintain accuracy.

Structured data delivery

Processed data was delivered in CSV and JSON formats for seamless integration into internal systems. Final documents were stored in standardized folders using predefined naming conventions to simplify downstream access.

Discrepancy and audit management

The system flagged missing or invalid documents and generated exception reports highlighting discrepancies. A robust audit trail was maintained to support future regulatory audits and compliance checks.

Results

• Automation reduced onboarding delays and sped up verification workflows.
• Verified data and documents were delivered in standardized CSV/JSON formats.
• Documents were automatically classified, stored hierarchically by PAN, and regularly updated.
• Outdated UBO data or missing consents were flagged instantly, enabling timely follow-ups.