Hey team, the Voice Lab transcripts have been missing for a few weeks now. I know this was flagged as a known issue and a patch was in the works, but it's still not working as of today. My client relies on the transcripts to review test calls and send feedback with screenshots. Without them, testing is significantly harder. Is there an ETA on the fix?

Hey team, running into a few issues with my plumbing AI receptionist and hoping to get some guidance. Tagging @Jorden Williams @Mike Copeland @Assistable Ai @Assistable Team 1. Character Limit (8017/8024) I'm basically maxed out on my prompt and need to add more instructions. Are there tricks for reducing character count while keeping the same behavior? Specifically: - Can I move longer scripts (like a company value proposition pitch) into the Knowledge Base and just reference it from the prompt? - Is there shorthand or syntax that GPT 4o understands more efficiently? - Any formatting tricks that save characters? 2. Number Pronunciation Same issue as @Justin Abee posted about. My assistant says "$90" instead of "$99" and reads addresses weird on first attempt (said "28 Fiore 7" instead of "2847"). Using Eleven Turbo V2.5 with Speech-Friendly Formatting ON. Tried spelling out numbers as words in the prompt (e.g. "ninety nine dollars") but it still sometimes gets it wrong. Any fixes? 3. Value Building / Step Compliance Even with explicit instructions like "STOP until you have first AND last name" and "After ALL fields, do VALUE BUILDING before pricing," GPT 4o still occasionally skips steps. The data collection is much better now but it still skips the company pitch before quoting fees. Any tips beyond uppercase emphasis and MUST/STOP language to improve compliance? 4. Voice Tone Shifting The voice occasionally shifts tone mid-call (almost sounds like a different person for a sentence). Using Amber King voice with Eleven Turbo V2.5. Is this an ElevenLabs known issue or is there a setting to stabilize it? Appreciate any help. This is for a live client deployment so getting these dialed in is critical. Thanks!

Running a plumbing AI receptionist with a 12-step data collection flow. GPT 4o follows the steps most of the time but occasionally skips collecting the last name or doesn't ask for issue details before moving on. The instructions are explicitly in the prompt with "STOP" and "MUST" language but it still sometimes skips. Has anyone found techniques to improve step compliance? Wondering if the Knowledge Base feature helps reinforce instructions or if there's a better approach.

Has anybody else been having an issue with this Can somebody help me out with this?

Hey guys would love some support on this issue I have AI assistant that is not working inside of the voice lab neither when I call the phone number this is for million dollar Plumbing Company would love some help ASAP thank you 😅