Over the past few months, Twilio hosted a series of hackathons focusing on innovative applications of Large Language Models (LLMs) when combined with Customer Data Platforms (CDPs). The goal was to empower developers to leverage these technologies to create more personalized customer interactions. Let's take a look at some highlights from our most recent LLM LGTM 👍Hackathon with Google!
Twilio + Google
A huge thank you to the Google team for partnering with us on this event and offering to give developers an exclusive peek at GCP’s newest Generative AI offerings after Google I/O. We gave attendees access to the PaLM 2 API so that they could get hands on with the technology while learning about it. Google was instrumental in igniting creativity and broadening our participants' understanding of the potential of AI technologies.
Ankita Sharda and Anto Joseph gave us an overview of the Google Cloud AI products for every persona, from ML expert to developer to businessperson. They showed us some exciting new features in VertexAI like access to new foundation models, including Codey for text to code. And we were excited to learn that Generative AI Studio, Model Garden, and PaLM 2 for Text and Chat are now in preview...so of course we used the PaLM 2 API for hacking!
What we hacked on
Developers had access to a mock set of customer data for an imaginary business (generated using AI!) and a specially-trained LLM chatbot that they used to understand more about their business’s pains and needs. We then tasked developers with a simple prompt: Using that first-party customer data from the Twilio Segment CDP and generative AI from the Google PaLM API, solve one of your business's customer engagement problems, improve user experiences, and drive more personalized interactions with customers.
And armed with the right tools and the right problem, the building began!
Over the course of 3 hours, developers built some pretty amazing solutions. Here, we'd like to spotlight the first and second place projects. Our Twilio judges, Lizzie Siegle and Sai Pragna Etikyala, said these projects stood out for their innovation and creative application of the Google PaLM 2 API and the Twilio Segment CDP data.
First Place: Shadazzle Answer-a-tron 5000
This team dazzled us with their solution to enhance customer service experiences with AI and customer data. Their prototype combined personalized responses, a voice assistant, and seamless integration with customer data.
Second Place: Team 2 | FitLyfe PaLMBot
FitLyfe PaLMBot, brilliantly leveraged AI to deliver a sophisticated fitness assistant. Their solution, characterized by an intuitive user interface, robust payment system integration, and a forward-thinking approach, embodied innovative design and execution.
Congratulations again to all our winners and participants!
If you’re excited by the potential of AI to improve customer experiences, join us for free and virtually on August 23, 2023 for SIGNAL, Twilio’s yearly customer and developer conference where we’ll explore the exciting world of Generative AI together and see how we can merge Large Language Models with solid, compliant data to create even better customer experiences. Register now!