Anyone switched to 5 Nano model yet? If so, curious how it runs. It's faster, cheaper and 2.5x the context window. I use 5 Nano in a lot of my other stuff (its awesome), but haven't tried it for chat/voice. If you've used it with Assistable, how did it do?

Have any questions? Text me (iMessages & SMS): 770-296-7271 Need white-gloved help RN? (I am in there too): help.assistable.ai Have anymore questions? Text me and I will make a video over: 770-296-7271 ----- where have i been? - Actively trying to fire myself. I think a long term worry internally, myself, and from the community is key man risk. My only objective is trying to make sure this outlives me. Im mad. - Me too. Lets figure it out together, reach me or team and lets make sure clients are super duper happy Why is X happening? - I explained in the vid along with solution - if you are super busy, text me and ill get you an answer in less than 5 mins What can I expect? - A platform that does rely on me or any vendors - something that we own in entirety and can be maintained by people much smarter than me X ? - Y A (aka text me or run it up to support and they will send to me) we are on call 24/7, ask anyone thats talked to me in the past week - I have slept collectively 6 hrs (not even in my bed :/ ), yelled at some people upstream, and gave every ounce of my time to help anyone down stream I miss making videos and being involved here. I want to do it more. Thats what we are trying to do here

Any meaningful upgrade comes with a moment of discomfort. V3 is one of those moments. To support what’s coming next, we had to migrate some core components earlier than planned. While this was the right decision for the long term, we know the timing has caused real frustration. If you’ve experienced call issues, delays, or unexpected behavior, we want to be clear from the start: We see it. We understand the impact. And we’re actively fixing it. Right now, there are three main things we want to acknowledge. First, there is a delay in the AI agent greeting on outbound calls, and this is affecting all outbound traffic. We know how important that first moment of a call is, and we agree that this experience is not where it should be. A fix for this is part of the upcoming patch and is being treated as a top priority. Second, some users are experiencing latency or connection issues. The majority of the platform is operating normally, but we already have a patch in progress that directly addresses these issues and will be rolling out shortly. Third, there is a busy signal issue affecting a small subset of numbers. We currently host over 14,000 phone numbers. Out of those, a few hundred are affected. If you’re one of them, your frustration is valid, even if the issue is isolated. Most of the impacted numbers fall into one of these categories: - Custom Telnyx setups that don’t align with our updated requirements - SIP trunking requests that were never completed - Older numbers still in the final stages of porting from Twilio to Telnyx If you’re affected, there are two paths forward. Option one: wait it out. For many numbers, this will be resolved automatically over the next few days as the patch and cleanup complete. Option two: move faster. You can purchase a new number and let our support team know. If your old number qualifies, we’ll credit your wallet for the entire year for the phone number ($30) for the trade-in. This option is there for anyone who can’t afford downtime.

Is there a list of potential values that would be inputted into each section of a webhook? For example, under "disconnection_reason" --> what are all the possible outcome, on top of "user_hangup"? "call_id": "call_712...ae", "call_type": "phone_call", "direction": "outbound", "to": "+17704537225", "from": "+14703865498", "contact_id": "hCCG...5va", "disconnection_reason": "user_hangup", "user_sentiment": "Neutral", "call_summary": "The user...e call politely.", "call_completion": "Complete", "call_completion_reason": "user_hangup", "assistant_task_completion": "Complete",

Been going pretty deep on vibe coding lately and learning a lot as I go. Started out using Lovable and spent way more money than I want to admit. And I kept fighting it because of the context window. I'd be going back and forth trying to get it to do what I wanted and it just kept losing track. That's when I started learning about how these tools actually work under the hood. Switched to Claude Code to save money and started figuring that out. One thing that kept bugging me was waiting for one thing to finish before I could start the next. When you're trying to actually build something out, that adds up fast. So somewhere in that process I ended up building a tool for myself. It lets me map out features on a Kanban board and have multiple AI agents work on them at the same time. Each one stays in its own git branch so nothing gets tangled up. It figures out what needs to happen first and keeps everything organized. Wasn't really planning to share it but I know a lot of people are in this same learning process right now. Figured I'd take the extra step to package it up and make it easy to download in case it helps someone else too. It's free if you want to check it out. agenticstudio.hudley.ai Let me know if it's useful. Happy to share more of what I've been working on 🤙