RICSA - AI Audit Chatbot

AI audit chatbot built on Retrieval Augmented Generation (RAG). It answers from regulations, policies, SOPs, and past audit findings, with every answer traced to its source.

Story

Auditors must examine evidence that is accurate, consistent, and regulation-based. They need access to internal and external regulations, company policies, SOPs, previous audit reports (LHA), and data from audit applications spread across different systems. Searching and tracing this information today is manual, slow, and open to different interpretations, which weakens findings and recommendations. RICSA answers this challenge: an AI Audit Chatbot based on Retrieval Augmented Generation (RAG) that retrieves audit information from source documents and shows where each answer came from. The system is designed as a decision support tool for auditors, not as an audit decision maker.

Contribution

• · Built the RAG-based AI chatbot chat interface for fast audit information retrieval.
• · Built the document library view to index regulations, policies, SOPs, and audit reports.
• · Built the audit trail view so every answer can be traced back to its source document.
• · Built the AI temperature / model settings view for flexible chatbot configuration.

Impact

Speeds up information search and understanding of regulations, policies, SOPs, and audit findings, increases auditor effectiveness and productivity, and keeps interpretation of regulations consistent across auditors.