N8n Central Bank Sentiment Tracker: OpenAI + Sheets + Slack

Get a 3-hour early warning when central bank language turns hawkish or dovish—then log it to Google Sheets and alert your Slack.

This n8n Central Bank Sentiment Tracker pulls fresh FED, ECB, and RBI statement data from an HTTP feed, uses OpenAI (GPT-4.1-nano) to score hawkish/neutral/dovish sentiment, writes structured results into Google Sheets, and posts a Slack alert when the latest sentiment shifts by 20+ points versus the rolling average.

What this workflow does

• Runs every three hours using an n8n schedule trigger.
• Fetches new statements from an HTTP endpoint, then normalizes/validates records (required fields, minimum length, supported bank codes).
• Deduplicates by loading previously processed statement IDs from the Google Sheets “Sentiment Log” sheet.
• Analyzes sentiment with OpenAI: sends each new statement to GPT-4.1-nano and requests JSON containing classification (hawkish/neutral/dovish), sentiment score, confidence, key phrase, summary, and policy signal.
• Stores results in Google Sheets by appending the validated structured record to the “Sentiment Log” sheet.
• Calculates a rolling sentiment shift vs. the previous five logged scores for the same bank, flagging shifts of 20 points or more.
• Notifies in Slack with a detailed message when meaningful shifts occur.

Use cases

• Monitor central bank policy tone for FX, rates, and macro teams.
• Automate internal “risk-on/risk-off” signals based on real-time hawkish/dovish wording.
• Give SaaS operators and automation engineers a reliable sentiment + alert pipeline with audit-ready logs.

Technical details

• Nodes/stack: HTTP Request, if, Code, OpenAI (GPT-4.1-nano), Google Sheets, Slack, Sticky Note.
• Data flow: HTTP feed → validate/normalize → dedupe via Sheets → OpenAI JSON parse/validation → append to “Sentiment Log” → rolling average comparison → Slack alert.
• Setup: provide OpenAI credentials and confirm model availability; connect Google Sheets; ensure the “Sentiment Log” sheet contains columns matching appended fields (e.g., statement_id, bank_code).