May 2025 Summaries

The tutorial outlines the process of building a voice-driven WhatsApp ticket booking app using AI, Twilio, and .NET. The application accepts WhatsApp voice notes or texts, uses AI to transcribe and interpret booking requests, checks availability, processes payments via a dummy service, assigns seats, and delivers PDF tickets by email. To evolve this application into a production-ready platform, consider integrating a real payment gateway, persisting conversation and booking data, and adding user authentication and profiles. The tutorial also provides additional resources for continued learning and related posts on Twilio Docs, Resource Center, Ahoy, and more.

Twilio's ConversationRelay service allows developers to integrate Speech-to-Text (STT), Text-to-Speech (TTS), and AI Large Language Models (LLMs) with their Twilio Voice Applications. Recently, ElevenLabs has been added as a TTS provider, offering over 1000 voices to choose from. This tutorial guides users through integrating and customizing ElevenLabs voices with ConversationRelay in Node.js. By following the steps outlined in the tutorial, developers can create AI chatbots or personalized assistants that respond to customer queries through natural voice interactions, without having to connect with an agent. The tutorial covers setting up a ConversationRelay service, importing environment variables, integrating ElevenLabs voices, customizing voice settings, and connecting the application to Twilio. Once completed, users can run and test their application, refining the voice settings as needed. By integrating ElevenLabs with ConversationRelay, developers can create rich, human-like voices that bring their voice applications to life.

The tutorial provides a step-by-step guide on how to zip files in Go programming language using the standard library. It covers setting up Dropbox and Twilio accounts, installing required dependencies, creating a main function that orchestrates the entire process, and serving as an HTML file used as the user interface. The application compresses uploaded files into a ZIP archive, uploads them to Dropbox, generates a temporary download link, and sends it to a predefined phone number via SMS using Twilio.

Twilio is partnering with OpenAI to enable developers to build communications applications using the Model Context Protocol (MCP) and the OpenAI Responses API. This integration allows for standardized context provision to large language models (LLMs), enabling seamless connections between MCP servers and tools accessible via the internet. With just a few lines of code, developers can leverage Twilio's capabilities through the OpenAI Responses API, deploying their own Twilio MCP server and experimenting with various use cases. Tutorials and repositories are available to help users get started, showcasing how to spin up a self-deployed Twilio MCP server and integrate it with the OpenAI Responses API. This partnership aims to unlock the full potential of this new technology for Twilio customers, enabling them to accelerate their development and build innovative communications applications.

Gmail is introducing a new subscription center that simplifies the process for users to manage and unsubscribe from unwanted marketing emails, enhancing user control and experience. This feature allows subscribers to sort their email subscriptions by sending frequency and unsubscribe from multiple brands simultaneously, without needing to open individual emails. The interface displays brands according to their email frequency, which can influence subscribers' decisions on whether to unsubscribe. This innovation is beneficial for email marketers as it can lead to higher open rates and fewer spam complaints by removing disengaged subscribers, ultimately improving sender reputation.

The tutorial outlines the process of building a customer review application using Golang, HTMX, Twilio, and Gemini. The application allows users to submit reviews, which are then analyzed by Gemini for sentiment analysis. The results are displayed on the webpage, and a notification is sent via SMS to the product owner with the rating submission. To build this application, developers need to have basic knowledge of Go and its concepts, as well as create an account with Twilio and set up their environment with necessary dependencies such as Gemini API key, Twilio credentials, and SQLite database. The tutorial covers how to initialize the Gemini AI client, create a function for sentiment analysis, send SMS notifications using Twilio, handle form submissions, and test the application.

The article highlights key regulatory updates for May 2025, focusing on telecom regulations and carrier changes. These updates may impact how businesses communicate with customers, especially when it comes to SMS and voice traffic. The main areas of focus include global trends such as the push for Sender ID registration, which is accelerating across countries like Czech Republic, Ireland, Qatar, Tanzania, and Brazil. In the U.S., the FCC's new STIR/SHAKEN rules expand the definition of "Voice Service Provider," requiring businesses to obtain their own SPC token and STIR/SHAKEN certificate. Additionally, new requirements for phone number purchases are being introduced in Brazil, Ireland, and Belgium, including KYC requirements and approved Regulatory Compliance Bundles. The article provides guidance on what businesses should do now, including auditing phone numbers and Sender IDs, registering early, checking with legal counsel, and using available tools.

SIGNAL 2025 was a two-day event that brought together over 850 developers, engineers, and builders from across the world to experience the future of customer engagement. The event focused on giving builders the how, showing off real code, and letting attendees experience Twilio's capabilities hands-on. Key announcements included the launch of ConversationRelay, a conversational AI capability designed for developers, and Conversational Intelligence, which provides automated analysis of voice calls and text-based conversations. Other notable launches included Messaging Channels Upgrades for RCS and WhatsApp, Event-Triggered Journeys, and advancements in platform trust, security, and developer productivity. The event also featured immersive experiences such as Twilio Arcade Activations that connected data with fun, including Owlvin AI, Owl Ops, Live Photobooth, and a live magic show powered by Twilio and AI.

This tutorial series is about integrating Twilio's ConversationRelay with OpenAI for real-time voice conversations. It covers setting up the environment, writing the server code, and adding tool calling to interact with external APIs. The tutorial assumes prior knowledge of Node.js and Twilio, and provides instructions on how to deploy the application using ngrok tunneling service. The series includes topics such as interruption handling, token streaming, and tool calling, which enable developers to build more interactive and engaging voice applications.

Dylan Frankcom is a software engineer with a passion for building tools that help people. He started playing games on computers as a child, and this sparked his curiosity about technology. As he grew older, Dylan became fascinated with the early years of Apple and Microsoft, and he was inspired by the ideas of Steve Jobs and Bill Gates. He began teaching himself programming skills and eventually landed a job at a small startup in San Francisco, where he worked as a developer advocate. After college, Frankcom made contributions to open-source projects and started teaching iOS development at his alma mater. He is now joining Twilio's Developer Voices team as a Python Software Engineer of Technical Content, where he will work alongside people who share his values and passion for building tools that empower others.

The Twilio Arcade is an immersive experience at the SIGNAL conference in San Francisco, featuring a retro-futuristic atmosphere powered by Twilio products and services. Attendees can engage with various experiences, including a live caricature booth where they can receive a personalized sketch via video chat, create their own conversational bot using Owlvin AI, play action-packed challenges with Owl Ops, and discover the secret backroom of the Twilion's arcade machine. These experiences are built on top of Twilio products such as Segment, Twilio Video, Programmable Messaging, and SendGrid, which provide data-driven magic, real-time video chats, notifications, and email delivery. The experience is designed to demonstrate personalized adventures that delight users, inspiring developers to create their own great experiences with Twilio's suite of products and services.

You can now set up a native integration with Conversational Intelligence and ConversationRelay using Node.js. This integration allows you to track the performance of your AI agents outside test environments, ensuring they act responsibly, meet user needs, and deliver business value. To deploy this tutorial, you need Node.js, Twilio's Voice capabilities, ngrok or another tunneling service, an OpenAI account and API key, and a phone to place outgoing calls to Twilio. You create an Intelligence Service with Conversational Intelligence, add language operators, build a virtual agent with ConversationRelay, and set up the server. The integration automatically transcribes conversations in your Conversational Intelligence Service while running language operators, providing insights into conversational performance. After setting up the integration, you can test it by making calls to your Twilio number, interacting with the AI assistant, and verifying that the transcript is stored in your Conversational Intelligence account. The integration also provides quantifiable answers to questions about the AI's behavior, such as hallucination, customer emotion, and task completion rates.

ConversationRelay is a cloud-based platform that enables developers to deliver human-like AI voice experiences with minimal infrastructure management. Launched in public beta in November 2024, it has gained significant momentum and is now generally available as a General Availability (GA) product. The platform provides flexible configuration options, blazing-fast audio, and low-latency interactions, making it suitable for various industries such as healthcare, retail, financial services, and enterprise customer experience. With its Conversational Intelligence integration, ConversationRelay offers real-time insights from every interaction to continuously improve voice AI performance. Its HIPAA eligibility makes it an attractive option for regulated industries, while its support for multiple Speech-to-Text (STT) and Text-to-Speech (TTS) engines expands its flexibility. By removing the burden of managing fragmented infrastructure, ConversationRelay empowers businesses to focus on intelligence rather than infrastructure, enabling them to build smarter, faster, and more personalized experiences.

At Twilio SIGNAL 2025, the company unveiled its latest innovations in customer experience, AI, and global communications. The event highlighted Twilio's strengths in CPaaS and CDP, enabling businesses to create personalized experiences with AI-powered solutions. Key announcements include Twilio's Conversational AI platform, which brings together human conversation and AI intelligence, and its new partnership with Microsoft to accelerate AI-powered solutions for customers. Additionally, Twilio strengthened its reliability features, including Customer Deliverability Engine and Fraud Guard, to improve message deliverability and protect against fraud. The company also showcased its Customer Data Platform, Twilio Segment, which empowers every team with a centralized view of customer data across the entire journey. With these updates, Twilio aims to empower businesses to build trusted, simple, and smart communications platforms that accelerate AI-powered solutions for customer experiences.

Conversational Intelligence is a unified AI understanding platform that blends AI, data, and omnichannel communication to create a seamless customer experience. It breaks down silos in customer engagement by unifying conversational data across voice, messaging, and virtual agents, giving businesses a true, cross-channel view of the customer journey. This enables smarter automation, faster resolution, and more connected experiences. The platform introduces powerful new features such as AI Agent Observability with ConversationRelay, Generative Custom Operators for language analysis, and Conversational Intelligence for Messaging. These features empower businesses to unlock unified AI understanding across Voice, Messaging, and Virtual Agents, driving continuous improvement and measurable impact. With the integration of Generative Custom Operators, businesses can automate call scoring based on what really matters, enhance agent performance, optimize sales cycles, and increase customer satisfaction at scale. Conversational Intelligence is now available in private beta, allowing businesses to request access and seize the opportunity in every customer conversation.

Time is of the essence for businesses seeking to deliver precise, scalable, and context-rich personalization in real-time. To address this challenge, Twilio Engage has been redesigned with a new Journeys architecture that unlocks highly personalized and transactional customer journeys for businesses. The new architecture introduces Event-Triggered Journeys, which use intent signals and event-triggered architecture to enroll members and make decisions. This approach enables the dynamic management of state in "cart" scenarios without requiring developer support, scaling to whatever business requirements and data are needed. Additionally, Journeys will create rich contextual payloads that combine information from triggering events, data stored in the warehouse, and customer profiles, allowing for personalized responses with all relevant information. The new architecture also includes improved observability and usability, enabling customers to visualize Journey effectiveness and performance in real-time.

Twilio has been named a leader in Communications Platform as a Service (CPaaS), Customer Data Platforms (CDP), and Customer Engagement Platforms (CEP) by Omdia Universe reports. The company's platform convergence strategy, which combines communications, data, and AI capabilities to provide unified customer experiences, is driving its success. Twilio's unique combination of CPaaS, CDP, and CEP capabilities sets it apart from competitors, with features such as AI-powered fraud prevention, conversational intelligence, and vendor-agnostic approach to data integration. The company's focus on emerging growth channels, user authentication, and AI-enabled use cases is also driving its leadership in the market. With a strong emphasis on customer outcomes, Twilio is well-positioned to lead the future of customer engagement, which will be characterized by unified experiences at scale and flexible API-first architectures.

The author of the post discusses how to enhance a Twilio Voice application built with Anthropic's Claude models by adding token streaming and interruption handling. The original setup had latency issues, particularly when dealing with long responses. Token streaming allows for real-time response delivery, reducing latency. To implement this, the `aiResponseStream` function is modified to handle streamed tokens and forward them to ConversationRelay. Additionally, conversation tracking and interruption handling are added by modifying the `handleInterrupt` function to replace interrupted parts of the AI's turn with what was heard. The updated application allows for improved responsiveness and better handling of interruptions during conversations.

You can integrate Claude, Anthropic's AI assistant with Twilio Voice using ConversationRelay. Twilio's ConversationRelay allows you to build real-time voice applications with AI Large Language Models (LLMs) that you can call using Twilio Voice. This tutorial demonstrates a basic integration of Claude with Twilio Voice using ConversationRelay. To deploy this tutorial, you need Node.js 23+, a Twilio account and phone number, an Anthropic account and API Key, Ngrok or another tunneling solution, and your code editor or IDE of choice. You create a project directory, initialize a new Node.js project, install the necessary packages, configure environment variables, write the server code, start ngrok and the server, configure Twilio, and test your integration. The code integrates Claude with Twilio Voice using ConversationRelay in under 100 lines of code, allowing you to build real-time voice applications with AI LLMs that can engage in conversation with an Anthropic model.

To create an interactive voice call survey system with Twilio and Go, you'll need to initialize a new Go project, create environment variables for the credentials, retrieve your Twilio credentials, install Twilio's Go Helper Library, install other dependencies, define the database schema, create the survey logic, create the application templates, make the application accessible on the internet, test the application, and finally serve as the entry point of the application. The system will allow you to add survey questions, invite customers to participate via phone call, view statistics on the collected responses, and make informed decisions based on customer preferences and feedback.

The tutorial demonstrates how to automate call center management using Go and Twilio's TaskRouter. It covers setting up a workspace, configuring workers, creating task queues, defining workflows, handling incoming calls, redirecting missed calls to voicemail, updating worker availability via SMS commands, and testing the application. The process involves initializing the Twilio configuration, defining HTTP routes, and starting the server. The tutorial provides a foundation for building a call management system with Go and Twilio TaskRouter, with potential enhancements including event logging, missed call management, and integrating with databases or advanced reporting tools.

In an evolving data-driven landscape, the Twilio Segment User Group highlighted the necessity of bridging the gap between marketing and data engineering teams to foster a robust data culture within organizations. This involves focusing on data quality and observability, achieving a unified view of customer profiles, and leveraging AI for personalized customer interactions. Modern technological advancements, such as serverless technology and real-time processing, have shifted companies from traditional batch processing to a more dynamic infrastructure, enabling instant access to actionable data. Tools like Twilio Segment and Snowflake help eliminate data silos, allowing marketers to access insights without technical dependencies and to personalize customer experiences effectively. Establishing a data culture requires not only investments in technology but also a collaborative mindset, with structured workflows, education, and governance to maintain data integrity and accessibility. By embracing a hybrid analytics model and fostering cross-team collaborations, organizations can enhance engagement and decision-making, leveraging customer data platforms to break down silos and improve data quality for scalable personalization.

You can play MP3 audio files during Twilio voice calls using the TwiML `<Play>` verb in Node.js. To do this, you'll need to set up a Node.js application with Twilio credentials, host an MP3 file either through Twilio Assets or your own Express server, and use the TwiML `<Play>` verb in your Twilio Programmable Voice API calls. You can play MP3 files on both incoming and outgoing voice calls using separate code paths, and you can also add additional functionality such as playing DTMF tones or looping audio files. The process involves setting up a local Express server to host the MP3 file, creating an ngrok tunnel to make the server publicly accessible, and then configuring your Twilio number to use the forwarding URL to play the MP3 file during incoming calls, or using a different code path for outgoing calls.