ad
ad

Developer Lab: Building an SMS and Voice AI Real Estate Agent

Science & Technology

Introduction

Welcome to our third-ever DevLab! We’re thrilled to announce that this is the largest event we've hosted, with over 500 registrations. Thank you for your continued support and enthusiasm!

Agenda

Today's workshop is dedicated to demonstrating how to build an SMS and Voice AI Real Estate Agent. Here’s a brief agenda:

  1. Quick introductions and a special announcement.
  2. A community demo from Paul.
  3. Workshop portions where we will:
    • Present a demo.
    • Conduct a voice walkthrough.
    • Explore SMS setup.
    • Wrap up with a Q&A session.

We encourage everyone to stay engaged throughout the workshop. Feel free to use the chat and Q&A tabs to ask questions at any time, and if you want to, you can raise your hand to ask your question live.

Introduction

I’m Kim, the community lead at Voiceflow, working on community programs and events. Now, I'll hand it over to Nico for his introduction.

Nico’s Introduction

Hi everyone! I’m a product advocate at Voiceflow, focusing on our AI agents and community growth. My role involves helping users discover new features and share resources with everyone.

Special Announcement

Now, let's make a big announcement! We're excited to announce the launch of Voice and Voiceflow in beta. This is a comprehensive suite of tools designed to help you build and deploy voice-enabled AI agents seamlessly.

Key Features of Voice in Voiceflow:

  • Low latency with high accuracy in interpreting user speech.
  • Multimodal capabilities for seamless integration with web chat experiences.
  • Tooling powered by industry leaders like Deepgram and 11 Labs, with integrations available for platforms such as Twilio and Vonage.

During today's workshop, we'll be using this beta, demonstrating how to build complex voice AI agents.

Workshop Focus: Real Estate Use Case

We chose real estate as our use case for today due to the industry's need for efficient lead generation and customer support; agents often juggle many tasks across platforms such as WhatsApp and Excel. Customers require quick responses, and the challenges of modern consumer tech platforms drive the necessity for AI agents capable of fast interactions.

Guest Speaker: Paul

Paul, who has been actively involved in developing a solution for a real estate client, Property Tree, joins us to share insights from his work.

Paul's Project Overview

Hi everyone! I’m the lead AI engineer at Fast Brand Technologies. We recently collaborated with Property Tree to address their challenges utilizing AI tools.

Our goal was to automate communication, streamline lead engagements, and ensure customer inquiries are addressed promptly, minimizing the burden on real estate agents.

The initial phase of our project aimed to capture leads through WhatsApp, allowing agents to focus on more crucial interactions. I was able to build the agent in just under two hours, which greatly emphasizes the efficiency of the technology we have access to.

Workshop Demo

We move forward to present a demo of how the AI agent interacts with users. The demo showcases voice interaction focusing on gathering user preferences for real estate listings.

Technical Breakdown

After the demo, we delve into the technicalities of building the AI agent:

  • Agent Configuration: Using a single Voiceflow agent for both SMS and voice capabilities.
  • Data Handling: Leveraging user IDs for consistent data flow across platforms.
  • LLM Integration: Implementing LLM for enhanced conversation modeling and property recommendation.

SMS Integration & Sending Listings

We also cover how SMS integration works with Twilio, allowing the agent to send personalized messages seamlessly. Users can request additional information or schedule property visits via SMS, providing a cohesive communication experience.

Q&A Session

The workshop concludes with a lively Q&A session, addressing various queries related to transcripts, security, language support, and human handoff procedures, all while emphasizing the importance of maintaining user privacy and consent.

Conclusion

We thank everyone for participating in this workshop! If you’re interested in trying the Voice and Voiceflow beta, please join our waitlist. Remember, you can also check out our documentation and community updates for more resources.

Keywords

AI agents, SMS integration, voice technology, real estate, lead generation, Voiceflow, Twilio, LLM, customer interactions, property recommendations.

FAQ

1. What is the purpose of the SMS and Voice AI Real Estate agent? The primary aim is to automate and enhance customer interactions in real estate, ensuring quick responses and efficient lead handling.

2. Can the AI agent handle multilingual conversations? Yes, the beta supports several languages, and more documentation will be provided on this.

3. How does security and compliance work with voice calls? It's crucial to inform users that conversations may be recorded and to obtain their consent for sharing information.

4. What are the limitations of the SMS integration? Usual limitations apply, such as character limits on SMS and the handling of message duration.

5. Will transcripts of calls and messages be accessible? Yes, transcripts will be stored and can be accessed similarly to chat-based agents.