Why it matters
Multi-agent orchestration is one of the hardest unsolved problems in applied AI right now, and hackathons like this surface real working patterns faster than papers do.
The tokenmaxxing angle
Agent swarms multiply token spend by design — every added agent is another set of calls; a hackathon judging swarm projects is a direct proxy for how builders are (or aren't) managing multi-agent cost blowup.
From the organizers
Hosted by community 'tokens&' with sponsor Guild AI and inference partner Pioneer; speaker Corbett Waddingham (Guild.ai Head of DevRel) with judges from Guild.ai's growth and engineering teams.