Ben Dunphy

Ben Dunphy, also published as Benjamin Dunphy, co-founded the AI Engineer conferences with swyx. His work included program curation and the in-person experience, alongside swyx’s content curation. At AI Engineer Summit 2023, he introduced Network’s LLM-powered attendee matching and announced plans to open source both the event application and its matching algorithm.[1][2]

In a March 2024 Latent Space conversation, Dunphy recalled that he and swyx began discussing an event after meeting at a hackathon swyx organized in February 2023. They decided the emerging AI engineering community warranted a conference and moved forward with the AI Engineer Summit.[1]

Dunphy argued that vendor neutrality helps an event serve a varied technical audience: organizers can concentrate on the quality of content, speakers, relevance and the participant experience. He supported putting competing companies on the same stage.[1]

Explaining the 2024 AI Engineer World’s Fair, Dunphy emphasized bringing experts, builders and practitioners together to exchange ideas and reflect. He saw extended face-to-face conversations as a way to slow the online news cycle, describing an “in person festival of smiles, handshakes, connections, and in depth conversations.”[1]

In his 2023 Network announcement, Dunphy introduced the Generative Matching Algorithm, explaining that it used LLMs to connect attendees with people who could help solve their stated problems. An attendee could describe a problem in their Network profile, making that request the basis for a connection. At the time, matching was limited to in-person attendees of that Summit; the application also provided the schedule, expo information and nearby dining recommendations.[2]

Dunphy announced that both the application and matching algorithm were being open sourced, with the goal of supporting attendees at other events. He credited lead engineer Simon Sturmer for the mobile application and Suizette Steller for the matching algorithm, as well as Descope for authentication and Supabase and pgvector for database infrastructure.[2]

Before the AI Engineer conferences, Dunphy organized events and training for frontend developers. He had been involved in that community since 2014, organizing meetups in San Francisco and New York and creating a company that trained engineers across the United States in React and the wider frontend ecosystem. His QCon San Francisco 2022 biography identified him as the founder of EventLoop, an events production company and software service for hybrid and remote industry events, the organizer of Reactathon and the original organizer of Jamstack Conf. At QCon, he hosted the track “Rethinking Our Relationship with the Frontend.”[3]

Writing as @benghamine in his January 2022 article “How to write an effective conference talk abstract”, Dunphy recalled reviewing more than 160 submissions for Reactathon 2020 and discussing the strongest with his committee. Using an abstract by David Khourshid, he recommended a sequence of a concise hook, a clear problem, an intriguing proposed solution and explicit learning outcomes. He also stressed that limited schedule space can force organizers to reject strong proposals, advising speakers to submit early and often.[4]