What's new in Weld - October 2025

Hello! 👋

Welcome to the October edition of the Weld Recap, your monthly summary of the newest product improvements, feature releases, and community updates. This month, we expanded connector coverage, enhanced system stability, and made it easier than ever to manage and monitor your data pipelines.

✨ New features & connectors

BigQuery ETL Source

You can now use BigQuery as an ETL source, not just a destination. This allows you to extract data directly from your BigQuery datasets and move it into other systems, giving you more flexibility in how you use Weld within your existing data stack. Read more →

Intercom

You can now sync call records from Intercom with the new “Calls” data stream, making it easier to analyze support and engagement data alongside other customer touchpoints.

TikTok Ads

Added support for GMV (Gross Merchandise Value) Max campaigns, including three new data streams for deeper performance tracking. Now you can measure campaign results directly from your warehouse and fine-tune your ad strategy with confidence.

Klaviyo

The Klaviyo connector just got a major upgrade, now supporting 11 new data streams, including catalog items, coupons, forms, and tags. This means richer marketing data and a more complete view of your customer activity in Weld.

MySQL

We’ve laid the groundwork for Change Data Capture (CDC), coming soon to MySQL databases. This sets the stage for real-time data replication and faster syncs across your MySQL sources.

Permissions

We’ve added a new way to identify required permissions for each connection, starting with ELT sources — helping you set up integrations faster and with fewer errors.

API Keys

Connections created via the API are now tracked, making it easier to see which integrations come from external systems and maintain visibility across automated setups.

Integration Entitlements

A new system now allows granting customers access to specific integrations — even if they’re not part of their current subscription plan.

🚀 Improvements & updates

We’ve continued to focus on reliability, monitoring, and smoother syncs across the platform.

System stability & monitoring

  • New monitoring service to automatically detect and handle stale or stuck real-time syncs (CDC)
  • Smarter retry handling for jobs that fail due to memory limits
  • Added lock-checking to the data warehouse scheduler to prevent overlapping jobs

Connector enhancements

  • Shopify: Clearer naming for order line items and improved metafield updates (only changed fields are synced)
  • HubSpot: Better error handling for duplicate IDs and improved stability when duplicates occur
  • Excel: Large worksheets are now automatically split into smaller chunks to avoid timeouts
  • Microsoft Ads: Added date range limits to stay aligned with the API’s data retention rules
  • BigQuery: Connections can now be edited, and cached clients improve performance
  • Klaviyo: Added tag sub-tables for more accurate data modeling
  • PostgreSQL: Added validation to ensure replication permissions are in place for real-time syncs (CDC)

General improvements

  • Optimized real-time syncs (CDC) by consolidating delete handling and timestamp extraction
  • You can now set custom sync intervals for individual data streams
  • Improved error logging across integrations for faster, clearer debugging

🛠️ Bug fixes

A few highlights from this month’s fixes and refinements:

  • Klaviyo: Fixed duplicate id fields and issues with tag sub-streams and flow reports.
  • Gorgias: Fixed incomplete ticket data fetching.
  • HubSpot: Resolved naming issues for tables starting with numbers.
  • Shopify: Improved order processing for high-volume accounts to avoid data loss.
  • CDC: Fixed stream startup statuses, job stopping logic, and conflicts with full syncs.
  • Facebook Ads: Corrected default handling for empty arrays.
  • General: Fixed fallback date logic, duplicate connection name validation, and false-positive timestamp logs.

💬 Community & events

This month wasn’t just about product improvements, we’ve also been busy building new ways to connect and learn together.

We’re on Reddit!

You can now connect with other Weld users and the data community in our new Weld Reddit community, a place to ask questions, share learnings, and swap best practices.

Building Reliable Data Foundations for Business Value: Live Event

Join us for our upcoming event focused on building reliable data foundations together with Steep, Pyne, and Minimal Data.

We’ll share insights from data leaders on reliability, observability, and scaling with confidence. Plus, we’ll be serving craft drinks from ToØl and plenty of 🍕 to keep the conversations flowing.

🗓️ Register here →

Customer Story: eComplete

Learn how eComplete uses Weld to unify data across platforms, improve visibility, and scale faster with reliable pipelines.

📖 Read the case →

🛋️ Until next time

That’s all for October! Thanks for following along, and as always, if you have feedback, questions, or ideas, we’d love to hear from you.

— The Weld Team 💙