Recent content by pbxgeek

Yes enter your email credentials under Default Settings -> Email. Fax service still uses the old email settings. We are changing it soon

Make sure you have these settings enabled. Flush cache for a good measure too

Yes, you can reset the password manually like this. https://www.fspbx.com/docs/configuration/users-and-groups/How-to-reset-Superadmin-password-or-create-a-new-Superadmin/

Are you aware that some yealink phones come with built-in wifi? We use those instead

Check Freeswitch logs. There might be something.

ASSEMBLYAI_WEBHOOK_HEADER=X-FSPBX-Webhook @pbxer

In my experience, they operate differently. Between Polycom, Grandstream, and Yealink, I had to find a set of settings that worked for each phone, and they weren't the same when it came to TLS and SRTP. You can at least test without encryption and see if the park works. Just to make sure it's...

You are partially correct. If you use SRTP, you also need to register the phone using TLS. That will dictate which port is used. I would check FS CLI too. It sounds like you may be able to see some info there related to this. I can assure you it works on Yealinks and TLS with SRTP as I have it...

Is your call park working on another phone connected to the same domain?

This works for all my yealinks. See the screenshot

yes that's the one.

I want to highlight an important aspect of any PBX: we optimize server resource usage. Every stage of call transcription is managed through our scalable queuing system, which leverages Laravel queues, Redis, and Horizon. This setup ensures there are no noticeable spikes in CPU or memory...

Hey everyone, We’ve just released a new Call Transcriptions module in FS PBX, bringing AI-powered transcription and summarization directly into your PBX — no external scripts, no manual uploads, no patchwork integrations. Here’s what this release includes AI Voice Transcriptions (AssemblyAI)...

I'm in the same boat as you, and I've concluded that Freeswitch documentation is often outdated or non-existent, so no LLM models can be appropriately trained to produce good output. It's a hit or miss, and there is a lot of hallucination happening.

We have committed a fix for this and assisted you on GitHub. Adding a note here in case someone else is having the same issue.