Agentic Engineering Hack with Google DeepMind

Google DeepMind appeared Saturday as a named partner on the Agentic Engineering Hack, a one-day New York event aimed at builders working on AI agents and related developer tools. NYC B2B listed the event on its May 20-28 startup calendar as running from 6:30 a.m. to 4:30 p.m. in New York, while Google DeepMind’s public Luma calendar showed the event on Saturday morning and marked it sold out. The event page described the format as a single-day hackathon built around “real problems” and demos for a room of working engineers. Other named backers on the event listing included Datadog, Nimble and ClickHouse, alongside Google DeepMind. ### Who was actually hosting the event? Luma listed the event as “Agentic Engineering Hack” and showed it was organized by tokens&, Alessandro Amenta, Jacopo Piazza, Datadog and one other partner. (nycb2b.org) Google DeepMind’s own Luma events calendar linked to the hack as an external event, tying the company to the program through its developer events presence rather than a standalone DeepMind-hosted page. (luma.com) GarysGuide’s event listing used the label “Agentic Engg Hack” and described it as a one-day gathering for New York builders to “hack on real problems” and present their work. That listing also identified the venue as to be announced, suggesting organizers were distributing some logistics closer to the event. ### Why was Google DeepMind attached to it? (luma.com) Google DeepMind’s name appeared in the sponsor lineup and on its Luma calendar, which says the page is run by the Google DeepMind Developer Experience Team. The most visible named DeepMind participant was Thor Schaeff, identified on the event pages as a developer experience engineer at Google DeepMind. GarysGuide and the Luma listing both named Schaeff as a speaker and judge. (garysguide.com) The event materials did not describe a product launch, funding announcement or formal DeepMind program tied to the hack; they presented the company as one of the technical backers and on-site participants. ### What were participants expected to build? The event description said teams would work on “real problems” and build agents or related systems with support from partner tooling. (luma.com) The sponsor blurbs were specific about the stack organizers wanted participants to think about: observability through Datadog, web data through Nimble and database infrastructure through ClickHouse, with DeepMind positioned as the AI research partner. (garysguide.com) CompeteHub’s mirrored description called it a one-day hackathon in New York where participants would tackle real-world problems, network with industry professionals and pitch to a judging panel. A GitHub repository created for the event described one entrant, “AgentSRE,” as an autonomous incident-investigation agent built with Datadog, ClickHouse and Nimble, offering a concrete example of the kind of agentic engineering project the format was designed to produce. (luma.com) ### Who else was on the speaker and judge list? Luma named Rushikesh Akhare of Luminai and Uriel Knorovich, co-founder and chief executive of Nimble, among the speakers. The judges list also included Adam Stevens of Nimble, Subodh Chaturvedi of Airbyte, Raymond Lin of Crosby and Lihong Wang of Freeport, alongside Schaeff and Akhare. Those names suggest the event was drawing from early-stage AI infrastructure and developer-tools companies rather than a single corporate host. (competehub.dev) The published materials, however, framed the day around hacking and demos, not recruiting or investment meetings. ### What did the schedule look like on Saturday? GarysGuide published a tentative schedule starting with doors at 9:30 a.m. and kickoff and hacking from 9:45 a.m. to 11:00 a.m. (luma.com) Lunch was listed for 1:30 p.m., demos for 4:30 p.m. to 5:00 p.m., and closing awards for 7:00 p.m. NYC B2B’s calendar showed a broader 6:30 a.m. to 4:30 p.m. block for the same event, reflecting the kind of timing differences that often appear between aggregate calendars and organizer pages. As of Saturday, the clearest next step for anyone following the event was to use the Luma registration page and NYC B2B’s calendar listing for updates, while Google DeepMind’s events calendar continued to show the hack on its public schedule. (luma.com) (garysguide.com)