n8n Research Report Automation: OpenAI & Gemini Fact-Check

Turn research questions into a fact-checked, cited report—fully automated in n8n

This n8n workflow takes a research request from an n8n form, uses OpenAI and Google Gemini to draft a cited report, then performs an OpenAI-based fact-check against the provided sources. The verified markdown report is emailed via Gmail and logged to Google Sheets for traceability.

What this workflow does

• Collects your request: Receives a research question, a depth level, and a destination email via an n8n form.
• Plans targeted searches: Uses OpenAI (gpt-5.6-luna) to break the question into 4–6 focused search queries.
• Finds and curates sources: Searches each query using Bright Data SERP API, then extracts, deduplicates, and caps the source URLs based on the selected depth.
• Builds a reading corpus: Fetches sources with Bright Data Web Unlocker as markdown, cleans/crops text, and compiles a compact corpus for generation.
• Drafts with inline citations: Uses Google Gemini (gemini-3.5-flash) to write a structured report using only the provided corpus with inline citations.
• Fact-checks every claim: Uses OpenAI (gpt-5.6-terra) to audit each factual claim against the cited source text, returning supported/unsupported/contradicted verdicts with quoted evidence, plus a confidence score and flagged claims.
• Delivers and logs results: Emails the verified markdown report via Gmail and appends run details to Google Sheets.

Use cases

• Automate “research brief” creation for SaaS operators who need fast, cited answers.
• Generate and verify internal knowledge reports (with evidence) for automation engineers.
• Produce depth-controlled research digests for customer-facing enablement docs.

Technical details

• Integrations: OpenAI, Google Gemini, Bright Data SERP API, Bright Data Web Unlocker, Gmail, Google Sheets.
• Workflow nodes used: set, code, gmail, limit, split out, aggregate.
• Bright Data setup: Create a Bright Data account, configure a SERP API zone and Web Unlocker zone, then add HTTP Header Auth with Authorization: Bearer <YOUR_KEY> for both Bright Data HTTP requests (configure zone names in the “Set Research Parameters” step).