PCI-compliant payment processing
Thanks for waiting, should be ready in 1 week.
Posts from every feed QUEST has found across the web — blogs, news, podcasts, channels and more.
Thanks for waiting, should be ready in 1 week.
Hey @snikolova Our preliminary analysis indicates that the issue originates from the model deployment on AWS Bedrock. The team is investigating this further, and we’ll keep you updated. Regarding your question, the model’s string response contains \u00000000, which is not valid unicode either. Thank You
Hey @pragmaticbusiness I’ve passed this on to the team to look. Thank You
Hey @juleselhelou Thanks for the information. I’ll check this with the team and get back to you once I have an update. Thank You
It would be great to see which transition was used to go from one node to the other node in the conversational flow agents.
call_dc07904d17d2c9daa1a1f5365af
Hi, i followed your suggestion but still didn’t work. can you check what happened please. call id: call_4b5257e7af56b516e4189a0399a
Hey @ezeokekemm016 We’ve received your email ticket and will review the details there. Our team will follow up with you through email. Thank You
Hey @plamen Retell does not filter assistant text by `phase`. Text labeled `commentary` is passed to TTS, just like text labeled `final_answer`. In your transfer example, the apology and “Please hold…” message would be sent for speech before the tool call is handled, subject to normal call interruptions. If a response contains both commentary and a final answer, both can be spoken in order. Retell suppresses some repeated text based on text matching, but partial repetition can still be spoken.…
Thank you so much for your help
Hey @ezeokekemm016 I’ve shared this with the relevant team for review. Thank You
This is the email address: Ezeokekemm016@gmail.com . Thank you
Hey @ezeokekemm016 Could you please share the email address you used to contact us? I’ll pass this information to the relevant team so they can review it and guide you further. Thank You
Hi Retell team, I’m the founder of Radius AI (https://radiusai.tech), an AI real-estate assistant built on Retell. I’m requesting manual KYC verification because I’m based in Nigeria, and Nigeria isn’t one of the countries Persona offers for ID verification, so I can’t finish the standard verification flow. About Radius AI Our users are people looking to rent or buy property. Radius AI finds listings for them and then calls the listing agent or property owner on the user’s behalf to: book a…
Hey @pragmaticbusiness Thanks for sharing the information. I’ll check this with the relevant team and get back to you once I have an update. Thank You
Hi Retell Support, We need help understanding the execution order of Post-Call Extraction and Post-call functions in our Retell Workflow. Based on our call logs, we noticed that Retell evaluates and skips our Calendly functions (check_calendly_availability and book_calendly_appointment) before triggering the call_analyzed webhook. Our concern is that the Post-Call Extraction results may not be ready when the Calendly function conditions are evaluated. Could you please clarify: Do Post-Call…
+1 from us. We run a healthcare receptionist agent that warm-transfers callers to a staff desk, and we lose the second half of the call the same way. Here’s what we do in the meantime. It isn’t a full recording, but it covers most of what we needed: **Log the transfer outcome as a post-call analysis field.** Answered live, voicemail or no answer. We roll it up into a daily report, so the business can see which transferred callers never reached a person and needs a callback. **Send the context…
We run a Retell agent in production for a multi-clinic healthcare client, so here’s what has actually caught problems for us. **Post-call analysis fields, not transcript parsing.** We define custom post-call analysis fields for the outcomes that matter (was the transfer answered, did the agent deliver the info), then roll them up into a daily report the client reads every morning. We started out using transcript regex. It broke silently the first time someone reworded the prompt. The analysis…
@ppetrov I’ve just followed up with the team on this and will get back to you as soon as I have an update. Thank You
Are there any updates on this?
Hey @jesse.betancourt You may be able to use your Retell Twilio number for WhatsApp integration. You would first need to enable A2P to receive SMS, then import the number into the WhatsApp Cloud API. There is also an option called Meta Coexistence, which allows the same number to remain active on the WhatsApp Business app while also connecting to the WhatsApp Cloud API. However, this depends on the WhatsApp integration partner you’re using, as it requires Embedded Signup and a few additional…
Hey @plamen Thanks for providing the details. I’ll check with the relevant team and get back to you once I have an update. Thank You
Hi Retell team, We have an internal LLM test harness that tries to test various cases of our Retell agents. Since Retell is using Response API for OpenAI invocations we were wondering how "phase": "commentary" messages are handled by Retell e.g. Are they spoken out loud i.e. passed to TTS or Ignored for TTS purposes and simply added to the conversation as context For example, before a tool call we notice that OpenAI also returns a commentary message - will the customer hear it? [ { "type":…
Hey @f.barbosaww you’d need to set up Custom Telephony with your own SIP trunk. More details here: International calling rates for Retell numbers Connect Retell voice agents to your telephony provider Use the Test SIP Connection button (sends OPTIONS, not a real call) before saving. Ensure your Brazilian carrier supports standard codecs (PCMU/PCMA/G.722), allows Retell’s IPs, and has no destination restrictions on +55. If calls still fail, gather Call IDs, Org ID, and a PCAP, then contact us at…
ultranon19: Two questions about conversation-flow transitions, for predicting and testing routing: Are prompt transition conditions evaluated in a separate LLM call after the user speaks, or in the same request that generates the reply (e.g. edges exposed as functions)? Which model and settings are used for that decision, and does it see the full call history or only the current node’s turns? Global nodes don’t seem to trigger from inside a component/subflow: a user saying “I want a human”…
I’ve run into this exact thing, and you can usually narrow it down pretty quickly. Here’s how I’d check it: 1. Listen to the dual-channel recording, not just the mixed one Retell gives you a multi-channel recording for each call (recording_multi_channel_url in the call object via API). The caller is on one channel and the agent is on the other. Open it in something like Audacity and look at the caller’s channel: Flat line: no audio reached Retell. That’s either a silent caller or one-way audio…
Hi Retell team, ▎ ▎ I want to use my own SIP trunk from a Brazilian carrier to make outbound AI calls to businesses in Brazil (+55). ▎ ▎ Before I buy the trunk, I need to know: ▎ ▎ 1. Is calling Brazil (+55) allowed on my account with custom SIP trunk? ▎ 2. Is there any country restriction or permission on my account that blocks it? If yes, please enable Brazil (BR). ▎ 3. Two other users (threads 3387 ("Retell BYO SIP trunk not sending INVITE to provider - dial no answer") and 3527 (Outbound…
Hi Zaki, Haoyang from Canlah AI (Singapore, GMT+8). We build Retell agents end to end: multi-prompt and Conversation Flow, function calling for slot lookup and booking, and post-call webhooks into n8n/Make or our own backend. Revenue share works for us, and we’re happy to start with the paid test build. I’ve sent you a DM with details.
I’m exploring a tool for agencies running voice agents for clients. No product is live yet, and I’m not collecting payments. The concept is scheduled test calls that check whether an agent still handles key scenarios, failure alerts, and a short per-client health report. For your agency, would this solve a problem worth roughly $49/month, depending on the number of agents and checks? What would it need to catch that your current tools don’t? If you already pay for a solution, I’d be interested…
Hi all I run Finura Labs, a UK-based AI agency that builds voice agents for law firms. I handle sales, client relationships, discovery and conversation design, and I’m looking for a Retell developer to be my long-term build partner on a revenue share basis. What we build Mainly inbound agents for law firms: Answering calls (after-hours or 24/7) Qualifying new enquiries and capturing details and a matter summary Booking consultations into Cal.com or Google Calendar Warm transfers for urgent…
