Pandium Product Updates June 2026

New Metrics API & Super User Metrics Dashboard

We’ve added a brand-new Metrics API and a Metrics Dashboard to give you a real-time pulse on your integrations. Track failed runs (broken out by integration errors vs. platform errors), integration statuses (published, unpublished, draft), and tenant statuses (active, paused, archived, and new tenants over time). Behind the scenes, we built out data aggregations so these insights load fast, even across large windows of history. To learn more, please review our documentation.

Standardized Dates & Times Everywhere

Time should never be ambiguous. We’ve standardized how dates and times are presented across the Admin Dashboard and Marketplace, showing every timestamp in your local time with a full, consistent date/time format and time zone. A new time zone picker in the Admin Dashboard lets you temporarily view times in a different zone when you need to. On the API side, all customer-facing endpoints now return timestamps in a single standardized ISO-8601 format.

Run Logs via the Public API

Your customers can now pull integration run logs directly through the V2 public API, no Admin Dashboard required. This makes it easier to surface run history in your own portal, automate monitoring, and build self-service troubleshooting into your product. For more information, visit our API reference.

New Connectors: Mambu and GitLab

We’ve expanded our connector library with support for Mambu, and GitLab. GitLab arrives with full webhook support, so you can trigger integrations off repository events in real time. As always, these connectors are ready to drop into your integrations today.

Webhook Improvements

Beyond new GitLab webhook support, we’ve made webhooks smarter across the board. Pandium now automatically unsubscribes from all webhooks when a tenant is archived, so you don’t leave stale subscriptions behind. We also updated our GoHighLevel webhooks ahead of their July 1, 2026 deadline to keep those integrations running without interruption.

Integration Tips

Before you ship a Release, check your HTTP client’s timeout and retry defaults, they’re rarely what you want.

Most clients ship with no timeout at all. One slow upstream API and your integration run hangs until the job is killed, burning time and hiding the real error. Retries are the subtler trap: the library can’t know which of your calls are safe to repeat, so it either retries nothing or retries everything. That’s your call to make, not the lib’s.

Quick checklist before you trust the defaults:

• Set an explicit timeout: both connect and read. A hung socket should fail fast, not stall the whole run.
• Know what auto-retries: many clients don’t retry on connection errors out of the box. Confirm it, don’t assume.
• Decide retry-safety per endpoint: most POSTs are actually idempotent, but “most” isn’t “all.” Check each write yourself and use idempotency keys where the API offers them, rather than trusting a blanket library default.
• Back off, don’t hammer: exponential backoff with jitter beats a tight retry loop that trips rate limits.

A hung run at 3am is a lot more expensive than five minutes setting a timeout today.

‍