if only i could go right back to the start. lots of people talk about the wins in their journey, but i feel as a community, we need to share losses too. to be frank, there are more losses than wins, we can all relate to them. just publishing wins, can be a dangerous game, we need to get around each other. it's a tough journey and much harder when people only share their successes and rub it in. tbh, that's why I'm making this post. I wish someone warned me of these things before i started my ai agency. It would've saved me time, money and a whole tonne of energy. But truth be told, even if someone did tell me: i'm not sure i would've listened. the hype of getting started and my stubbornness might've stopped me but i've found regardless, the best way to learn is to experience the lessons yourself. especially young and hungry dudes. embrace both the ups and downs - understand it's all part of the process. If you embrace your challenges and help people avoid them, this community will be a better place. That's why I'm telling you two things: -actually learn from my mistakes -actually try to apply these lessons to make your life easier i touched on the above in my recent YT video. always here to help gang. both with general biz advice and personal advice. best, Jack :)

🎯 Looking for 3 businesses to transform their phone calls with AI (FREE setup) Hey everyone 👋 I run Keel, an AI voice agent agency that's helped dealerships, clinics, salons, and other appointment-based businesses automate their phone systems without losing the human touch. I'm opening up 3 spots for businesses that want to test-drive this technology: ✅ $0 setup fee ✅ Pay only for usage (minutes used) ✅ 100% transparent dashboard - you see every call, every cost ✅ Production-ready system with full testing before launch All I ask in return is a detailed case study once we get results. Who this is perfect for: - Businesses losing leads to missed calls - Teams spending hours on repetitive appointment booking - Companies wanting to screen/qualify leads 24/7 - Anyone tired of "please leave a message after the beep"

Any recommendations to buy European phone numbers? I am from Spain so if I want to use Voice Agents through phone calls I would get charged for international calls unless the numbers are spanish (both European and US numbers). I have tried purchasing spanish phone numbers in different providers and the experience is horrible. Some of them like Twilio or Plivo don't even offer Spanish phone numbers. Telnyx does offer them but you have to go through a painful approval process that requires you to be a registered business and provide a lot of documentation. Do you know any better alternative to be able to call Pipecat agents through a phone number and not get charged international call fees?

Hi everyone, I'm building a specialized voice assistant using **Pipecat Flows v0.0.22** and running into a frustrating issue with phone number collection that I can't seem to solve. ### The Stack - **Framework:** Pipecat Flows v0.0.22 (Python) - **STT:** Deepgram Nova-3 (Polish `pl`) - **TTS:** Cartesia (Polish voice) - **Transport:** Local WebRTC (browser-based, no telephony yet) ### The Problem When I dictate a 9-digit Polish phone number (e.g., "690807057"), the assistant receives partial fragments and processes them individually instead of waiting for the full number. For example, if I say "690... 807... 057" (with natural pauses), the bot splits it into: 1. "6" -> sent to LLM -> LLM complains "Received only 1 digit" 2. "980" -> sent to LLM -> LLM complains 3. "5" ... and so on. ### What I Have Tried I've gone through the documentation and tried several fixes, but the "defragmentation" issue persists. 1. **Deepgram Configuration (Current Setup):** I've configured the `LiveOptions` to handle phone numbers and utterance endings explicitly: ```python options = LiveOptions( model="nova-3", language="pl", smart_format=True, # Enabled numerals=True, # Enabled utterance_end_ms=1000, # Set to 1000ms to force waiting interim_results=True # Required for utterance_end_ms ) ``` *Result:* Even with `utterance_end_ms=1000`, Deepgram seems to finalize the results too early during the digit pauses. 2. **VAD Tuning:** - I tried increasing Pipecat's VAD `stop_secs` to `2.0s`. - *Result:* This caused massive latency (2s delay on every response) and didn't solve the valid STT fragmentation (Deepgram still finalized early). I've reverted to `0.5s` (and `0.2s` for barge-in) as `stop_secs=2.0s` is considered an anti-pattern for conversational flows. 3. **Prompt Engineering (Aggressive):** - I instructed the LLM to "call the function IMMEDIATELY with whatever fragments you have". - *Result:* This led to early failures where the LLM would call `capture_phone("6")`, which would fail validation (requires 9 digits), causing the bot to reject the input before the user finished speaking.

Most people want money but don’t want discipline. That’s why they stay stuck