Server Logs AI Visibility Integration

Finseo analyzes server access logs in its Bot Traffic dashboard. Upload a .log or .txt file or push NDJSON from your server, and Finseo keeps only requests from known crawlers such as GPTBot, ClaudeBot, PerplexityBot, and Google-Extended, so you see which pages the AI engines behind ChatGPT, Claude, and Perplexity actually fetch.

Finseo detects the format from the first lines of the file. It reads common nginx and Apache access log layouts with client IP, timestamp, request line, status, response size, referrer, and User-Agent, plus a space-separated CDN-style layout. For other formats, push NDJSON through the Server Logs / API connection instead, using the documented field names.

Finseo accepts .log and .txt files up to 1 GB per upload. Files up to 50 MB are processed immediately and the dialog shows the bot visits found and duplicates skipped. Larger files are queued and processed in the background, so you can keep working while Finseo filters the crawler requests out of the log.

No. Finseo parses each line, identifies whether the request came from a known AI or search crawler, and stores only those crawler visits. Lines from regular browsers are discarded during processing. For the NDJSON push, the setup dialog also recommends filtering to crawler requests before sending, which keeps batches small.

Nothing is counted twice. Finseo deduplicates bot visits by crawler, timestamp, IP address, and page path within a project. That applies across repeated uploads and across sources, so a log you upload for a period already covered by a Cloudflare, Akamai, or NDJSON feed only adds visits that are missing.

Open Bot Traffic, Sync, and click Connect on the Server Logs / API card. Finseo shows an endpoint and a Bearer token. Push NDJSON batches with curl or from your own tracking, plaintext or gzip. For nginx, the dialog provides a bot-only log_format and a cron job that rotates and pushes the file every 10 minutes.

Each line needs ClientIP, ClientRequestUserAgent, ClientRequestPath, and EdgeStartTimestamp. Optional fields such as host, method, full URI, referrer, status, response bytes, country, city, and time to first byte enrich the Crawled Pages and Performance views. Send the real client IP rather than your load balancer, because Finseo checks crawler IPs against published ranges.

AI crawlers like GPTBot, ClaudeBot, and PerplexityBot do not execute JavaScript, so tag-based analytics records none of their visits. The server access log records every request regardless. Finseo turns that log into per-crawler, per-page data, giving Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) decisions a factual basis.

An AI engine can only cite a page its crawlers can read. Bot Traffic shows which crawlers reach which pages, sorted by purpose (live answer, search index, training), and where they are blocked or hit errors, so you can fix access before it costs you visibility in ChatGPT, Claude or Perplexity.