Get quick hiring insights from public employer reviews without reading long pages. The flow gathers review data through Bright Data and turns it into a clear summary with Google Gemini. HR and employer brand teams can share fast takeaways with leaders and recruiters.

A manual start kicks off a Bright Data dataset run and returns a snapshot id. The flow checks the snapshot status in a loop, waits 30 seconds if it is not ready, and tries again until data is complete. When ready, it downloads the JSON. The text is split into smaller parts and loaded into the AI summarization chain. Google Gemini creates a short, readable brief. The summary is then sent to your webhook endpoint for easy handoff to reports or dashboards.

You need a Bright Data account with an API key, a Google Gemini API key, and a webhook URL to receive the result. Swap the sample company page with your target pages to match your hiring goals. Expect less manual reading, faster reporting, and consistent summaries for weekly talent updates. Good fits include employer brand tracking, recruiting campaign reviews, and quick briefing notes for HR partners.

• Manual start to run the review check on demand
• HTTP call to Bright Data to trigger a dataset snapshot for target URLs
• Set node stores the snapshot id for later requests
• Polling loop with status check, If condition, and a 30 second wait until data is ready
• HTTP download of the finished snapshot as JSON
• Text splitter with 100 overlap prepares long content for better AI results
• Summarization chain uses the Google Gemini chat model to create a brief
• HTTP call posts the final summary to your webhook endpoint

• Reduce manual review reading from 2 hours to 5 minutes per company
• Automate about 80 percent of data collection and summarizing
• Improve consistency of reports with one standard summary format
• Connect Bright Data, Google Gemini, and your webhook in one flow
• Scale to more pages with a reliable status polling pattern

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with Bright Data, Google Gemini and Webhook.site. See the Tools Required section above for links to create accounts with these services.
3. In your Bright Data dashboard, create an API key for the dataset API. In n8n, double click the HTTP Request to Bright Data node, choose the 'Credential to connect with' dropdown, click 'Create new credential', then follow the on screen instructions to integrate that service.
4. Open the HTTP Request to Bright Data node and replace the sample URL list with your target company pages. You can add multiple objects in the JSON array to crawl more pages.
5. Confirm the Set Snapshot Id node maps snapshot_id. No change is needed if the expression already reads from the trigger response.
6. Open the Check Snapshot Status node and select the same Bright Data credential. Make sure the URL uses the snapshot_id expression so each run checks the correct job.
7. Review the Wait for 30 seconds node. Keep 30 seconds for most runs. Increase it if your datasets take longer to build.
8. In your Google AI account, create a Gemini API key. In n8n, open the Google Gemini Chat Model node, click 'Create new credential', and follow the on screen steps. Keep the provided model or choose another available model if needed.
9. Check the Summarization of Glassdoor Response node. Ensure it receives the JSON from the download node and is linked to the Default Data Loader and the Google Gemini model.
10. Open the Configure Webhook Notification node and replace the sample URL with your own webhook endpoint. If you need a quick test URL, create one in Webhook.site.
11. Run a test with the manual trigger. Watch the execution: a snapshot_id should appear, status checks should cycle, the JSON should download, and a summary should post to your webhook.
12. Troubleshoot common issues: if you see 401 errors, recheck the Bright Data credential and API key. If the job never completes, increase the wait time and verify your Bright Data dataset is configured to fetch the target site. If downloads time out, raise the request timeout in the download node. For very long pages, adjust the text splitter size or overlap to improve AI output.