I'm building a voice AI real estate agent using Retell AI that searches properties via Make.com and Airtable. The flow works partially, but I'm getting empty results despite having matching data. The Flow Retell AI voice agent calls find_properties function Make.com webhook receives the parameters Airtable searches properties with dynamic formula Array aggregator collects the results Webhook response sends data back to Retell AI My Retell AI Prompt The agent "Lena" asks users about: (Translated from German) Location (city) District (optional) Property type (apartment, house, etc.) Number of rooms Budget If users say "doesn't matter", the agent suggests defaults like: "I'll search for apartments in the city center with at least 2 rooms up to €500,000. Is that okay?" Then calls find_properties with concrete values. Function Parameters json{ "ort": "Straubing", // City "stadtteil": "Zentrum", // District "immobilientyp": "Wohnung", // Property type "min_zimmer": 2, // Min rooms "max_preis_eur": 500000 // Max price } Make.com Setup Module 1: Webhooks (Custom webhook) Module 2: Airtable (Search Records) Module 6: Array aggregator Module 9: Webhooks (Webhook response) Airtable Database I have 15 properties with fields: Ort (Straubing, Regensburg) Stadtteil (Zentrum, Ittling, Nord, etc.) Art der Behausung (Wohnung, Einfamilienhaus, etc.) Zimmer (1-5) Preis (€) (167,310 - 1,152,300) My Airtable Formula AND( {Ort} = "{{1.args.ort}}", {Stadtteil} = "{{1.args.stadtteil}}", {Art der Behausung} = "{{1.args.immobilientyp}}", {Zimmer} >= {{1.args.min_zimmer}}, {Preis (€)} <= {{1.args.max_preis_eur}} ) The Problem Expected: Should find 4 matching properties for "Straubing + Zentrum + Wohnung + ≥2 rooms + ≤€500,000" Actual Results: Airtable Output: Total number of bundles: 0 Array Aggregator Input: All fields show *empty* Final Response: Returns null values to Retell AI What I've Tested ✅ Simple formula works: {Ort} = "Straubing" returns 10 bundles

Hey everyone!! I’m offering FREE AI-powered solutions – things like voice agents, smart automations, and custom AI tools – in exchange for a short video testimonial. Over the past few months, I’ve built several AI systems, including one for a marketing agency that’s now pulling in $5K+ in monthly recurring revenue and helping their clients save big on ops costs. I’m now looking to work with a few more folks who are curious about using AI in their business or workflows. Whether it’s saving time, improving customer experience, or just testing out what’s possible – I’ve got you. No catch – if you like the results, say so. If you don’t, be honest. That’s all I ask. 📩 Interested? Comment below or DM me – let’s build something awesome together!

Hi guys, I need an honest opinion. I want to build an outbound caller that reminds my dental client's patients of their appointments. I have played around with Vapi some months ago, but I also know that Retell and Synthflow are strong in that field. In terms of functionality it should be as follows: 1. Call and check if this showing up to the dental appointment 2. If yes - hang up 3. If no - ask if he wants to reschedule 4. If yes - make a note in a spreadsheet that he will be called back from a (human) colleague later 5. Mark as "called" in the patient list Which one has proven to be a great fit for your clients? Maybe you have built something similar already. Cheers, Stephan

Let me know the best User-Friendly tools for developing a WhatsApp Chatbot with Integration with Openai + Google Sheets for saving the conversation.

Hi all, I am trying to connect my vector shift Knowledge base to my Vapi caller but somehow I can't get the set up right. Can anyone please guide me on how to set this up correctly within Vapi? Thanks and much appreciated. Best regards, Stephan