Faster Website Training, One-Click Start

We're excited to upgrade Website Training so you can start training right away, no more waiting for long scraping jobs before you hit Train.

⚡ Key Improvements

• Instant URL estimate - Get an initial full-URL estimate in ~10 seconds.
• One-click start - Submit URLs immediately and let scraping run inside training.
• Async API training - API training returns immediately with scraping handled in the background.

🔍 Better Visibility

• View trained content - Click the eye icon to review content after training completes.
• Clear failure reasons - See if training failed due to manual stop, 403 blocking, or duplicate content.
• Status filters - Quickly filter by failed, training, or trained websites.

🧭 Quick URL Estimate

• Fast URL discovery - Get an initial list of full URLs in about 10 seconds.
• Ready to submit - Start training immediately without waiting for scraping to finish.

🧵 Train Websites in One Flow

• One-button training - Submit the training job once and process websites in sequence.
• No pre-scrape wait - Scraping runs inside the training pipeline.

👁️ View Scraped Content

• Review results - See the content that was successfully trained.
• Understand failures - Check reasons like manual stop, 403 blocking, or duplicate content.

🧠 How It Works

• Unified pipeline - Scraping and embeddings now happen in the same training job.
• Tradeoff explained - Training may take longer, but you avoid the pre-scrape wait entirely.
• No retrain change - The retrain flow remains the same.

Get started from your chatbot's Website Training screen and hit Train to run everything in one step.