High-Performance Processing and Improved Support Tools

This update significantly increases concurrent scraping capacity and reduces job failure rates for high-volume tasks. We have also improved API connection reliability and resolved inconsistencies in billing notification displays.

Highlights

Edit and delete support replies

Admins can now modify or remove their replies on support tickets directly from the dashboard. This allows for quick corrections of typos or the redaction of sensitive information without needing manual intervention.

Concurrent media processing

Media uploads and downloads, including snapshots and document generation, are now handled asynchronously. This prevents individual file operations from stalling other tasks, significantly increasing throughput for batch scraping jobs.

Efficient cache management

New distributed locking prevents multiple identical requests from triggering redundant data fetches simultaneously. This reduces backend load and ensures faster response times during traffic spikes for frequently accessed resources.

New Features

Edit and delete support replies

Improvements

Concurrent media processing

Efficient cache management

Increased Concurrent Scraping Capacity

We have optimized resource allocation to support higher concurrency for parallel scraping jobs. This improvement has successfully reduced intermittent job failures by approximately 29% for heavy workloads.

Bug Fixes

Accurate error reporting

The usage dashboard now correctly includes network-level failures, such as timeouts and DNS issues, when filtering by errors. You can now see a complete picture of job health, including failures that occur before a status code is assigned.

Reliable media exports

Fixed an issue where screenshot and PDF generation tasks would occasionally return errors. All generated assets are now consistently accessible across our entire processing network.

Personalized account alerts

Account re-engagement and credit exhaustion notifications have been restored and updated with personalized content. You will now receive timely, branded alerts regarding your account status and usage thresholds.

Visual spending limit tracking

Fixed a UI bug where account spending limits would appear unset immediately after being saved. The dashboard now accurately reflects your current limits and ensures they persist correctly across page reloads.

Enhanced extraction availability

Advanced content extraction features are now fully operational across all processing nodes. This ensures consistent, high-quality data parsing regardless of which part of our network handles your request.

System stability hardening

Hardened our infrastructure against connectivity and memory pressure during high-concurrency tasks. These changes prevent service degradation during peak scraping loads, ensuring your requests are processed without interruption.

General bug fixes and improvements

Various minor updates have been applied to improve dashboard synchronization, billing refund reliability, and extraction fallback accuracy.

Improved API Connection Reliability

We have resolved an underlying issue that could cause intermittent service availability problems. You will now experience more stable connections and fewer request timeouts during high-traffic periods.

Seamless coordinator-only setup

You can now successfully initialize environments that focus on centralized coordination without local processing. This fix ensures that specialized configurations pass all startup checks without manual intervention.

Enhanced API Connection Reliability

Backend updates now prevent specific connection errors that could occur during high-traffic periods or system deployments. This ensures more consistent uptime and fewer unexpected interruptions for your integrations.

Corrected Credit Balance Notifications

We resolved a display bug where automated onboarding and inactivity emails could show a negative credit balance. Your email notifications will now accurately reflect your available account credits at all times.

Security

Alert data integrity

Resolved an issue where special characters in system labels could result in malformed data payloads. This ensures that all automated alerts and notifications are delivered with valid, readable information.

Plus 37 internal changes for stability and performance.