Smart search enginePerplexityRecently announced withYelpPerplexity is working with Yelp to provide real-time restaurant recommendations for its chatbot. Aravind Sriniva, CEO of Perplexity, said that more and more people are using chatbots as regular search engines, so it is a reasonable choice to provide information such as restaurants directly from the source. To this end, Perplexity integrates Yelp's maps, reviews and other details to respond to users' inquiries for restaurant recommendations.
Srinivas said: "We were originally a tool that attracted people to do fact-checking and research. But research can involve anything, such as if you are a coffee lover, you want to know where there are good coffee shops." Previously, David Pierce of The Verge pointed out that how Perplexity mixes text and links when answering questions has become one of its selling points relative to other chatbots such as Copilot, Gemini or ChatGPT. Srinivas said that this was also one of the considerations for the company to eventually choose Yelp, and Yelp's data enables Perplexity to provide answers in a variety of ways, including adding maps, photos, and even citing reviews.
Perplexity’s response included a link to the business’s Yelp page so users can view more reviews or other details. The company did not disclose financial details of the deal and said it would not use Yelp’s data to train any models because it uses existing models such as GPT and Claude2 as the basis for its chatbots.
The partnership is part of Yelp’s Fusion program, which licenses data to third parties including BMW. “By integrating our highly trusted and rich local data into Perplexity, we are able to provide reliable local search results to their users,” Chad Richard, Yelp’s vice president of business and corporate development, said in an email to The Verge.
Perplexity isup to dateA company that works with data providers to obtain continuously updated informationAI CompaniesSrinivas noted that Perplexity is not working with Yelp, but is licensing its data. Google is reportedly paying Reddit $60 million per year for real-time data and using Reddit’s data to train its AI models. OpenAI has signed deals with news publishers such as Axel Springer to get training data.
While other chatbots can also recommend cafes and restaurants, much of that data comes from information models learned from the internet, almost as if it were searched for the user. Some chatbots benefit from services that already aggregate information about businesses. For example, Google's Gemini can provide a list of restaurants with opening hours, menus, and maps when prompted because it can pull information from Google Maps.
Perplexity plans to do more integrations like this in the future, and has already partnered with answer engine WolframAlpha for math calculations. Srinivas offered some potential examples, such as working with vendors that provide shopping data or financial data.
Notably, Perplexity’s Yelp integration is also available on mobile.