Chat-Based P&L & Balance Sheet Analysis with GPT-4 & PostgreSQL

This workflow empowers finance teams, accountants, and data analysts to query Profit & Loss (P&L) and Balance Sheet data interactively via chat. A chat trigger initiates the process, and an AI agent intelligently routes questions to the appropriate PostgreSQL database: P&L data from financial_agent_pl_reports for revenue, costs, expenses, or profits; Balance Sheet data from financial_agent_balancesheets for assets, liabilities, or equity.

Powered by GPT-4.1-nano, the workflow generates precise SQL queries, analyzes results, and delivers responses in clean, readable tables with two-decimal precision. A memory buffer maintains conversation context for seamless, multi-turn interactions, enhancing usability without manual database switching.

Benefits include time savings on ad-hoc analysis, reduced errors from manual querying, and accessible insights for non-technical users. Ideal for financial reporting, auditing, trend analysis, and decision-making in SMEs or enterprises with structured financial data in Postgres. Setup requires database credentials and table population for immediate deployment.