Twilio voice AI builds Guinness price index across pubs

AI Engineer Matt Cortland used Twilio voice technology to build the Guinness Price Index, which verified 6,554 Guinness prices across pubs and venues in the UK and Ireland.

Using Twilio Programmable Voice integrated with ElevenLabs, Cortland made thousands of calls to pubs to gather current prices for a pint of Guinness. The result was a nationwide index built from direct phone contact rather than online submissions or manual surveys.

The project began as an effort to track the changing price of Guinness, but it has also drawn attention as an example of how voice-based AI systems can be used for large-scale information gathering. By automating calls and collecting responses in real time, the system handled a volume of outreach that would have been difficult to match through traditional research methods.

The work also points to a broader shift in how AI tools are being applied. Much of the recent public focus has centred on text chatbots, but the Guinness index shows how voice systems can take on routine real-world tasks such as surveying businesses, checking prices and compiling live datasets.

Beyond chatbots

Telephone calls are notable here because pubs and bars often do not publish up-to-date pricing online, and prices can vary sharply by location. A voice system that rings venues directly offers a way to reach fragmented businesses and gather data from the source.

The project combined programmable communications with conversational AI to carry out calls at scale. ElevenLabs provided the voice AI, while Twilio supplied the calling infrastructure used to place and manage the outreach.

For developers and technology companies, the significance lies less in the subject matter than in the method. A light-hearted consumer tracker has become a case study in using voice AI for practical field research, where the target information sits offline and must be collected through direct interaction.

"What started as a project about the price of a pint became a real-world test of where voice AI is heading," said Matt Cortland, AI Engineer and Creator of Guinndex.

"Using Twilio's Programmable Voice infrastructure alongside conversational AI technology, we were able to scale a single idea into a country-wide index. What's exciting is that this points to a much bigger shift. AI agents are moving beyond chatbots and into real-world interactions. Platforms like Twilio will play a critical role in enabling those experiences at scale," Cortland said.

Research at scale

The project illustrates one of the clearest commercial uses for voice AI: repeated, structured contact with a large number of organisations. Rather than replacing customer service staff, the model here is closer to automated fieldwork, with software making thousands of similar calls to collect a single type of information.

That could have applications beyond drinks pricing. Retail checks, availability monitoring, booking verification and local market surveys are all areas where information may still be locked inside phone conversations rather than websites or digital feeds.

Twilio framed the index as evidence that developers are moving from experimentation to more practical deployments. The company, which provides communications tools used by businesses and software teams, has sought to position voice as a key area of AI adoption alongside messaging and other digital channels.

Peter Bell, Vice President of Marketing, EMEA at Twilio, said the Guinness index showed how developers were applying AI tools in day-to-day use cases.

"Developers are moving beyond mere experimentation with AI; they are now building practical, real-world applications that deliver tangible value," said Peter Bell, Vice President of Marketing, EMEA at Twilio.

"Guinndex is a compelling example of the Twilio platform in action. By seamlessly combining programmable voice, conversational AI and communications APIs, we provide developers with the essential toolkit to create experiences that are both highly innovative and deeply human," Bell said.

The dataset also touches on a consumer issue with broad public appeal. The price of a pint has become a marker of cost-of-living pressures, and Guinness in particular has gained unusual cultural prominence in recent years, making price comparisons both novel and practically useful.

By verifying thousands of prices through direct calls, the index offers a picture of regional and venue-level variation that would be difficult to compile by hand. More broadly, it shows how AI voice systems are being tested not only in customer-facing roles but also as research tools built to gather information from the analogue parts of the economy.